<%BANNER%>

Determination of Test Methods for the Prediction of the Behavior of Mass Concrete

Permanent Link: http://ufdc.ufl.edu/UFE0041062/00001

Material Information

Title: Determination of Test Methods for the Prediction of the Behavior of Mass Concrete
Physical Description: 1 online resource (277 p.)
Language: english
Creator: Ferraro, Christopher
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: calorimetry, concrete, heat, hydration, isothermal, mass, modulus, semi, strength
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Hydration at early ages results from chemical and physical processes that take place between Portland cement and water, and is an exothermic process. The resultant heat evolution and temperature rise for massive concrete placements can be so great that the temperature differentials between the internal concrete core and outer concrete stratum can cause cracking due to thermal gradients. Accurate prediction of temperature distribution and stresses in mass concrete is needed to determine if a given concrete mixture design may have problems in the field, so that adjustments to the design can be made prior to its use. This research examines calorimetric, strength, and physical testing methods in an effort to predict the thermal and physical behavior of mass concrete. Four groups of concrete mixture types containing different cementitious materials are examined. One group contains Portland cement, while the other three groups incorporate large replacements of supplementary cementitious materials: granulated blast furnace slag, fly ash, and a ternary blend (combining Portland cement, fly ash, and slag).
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Christopher Ferraro.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Tia, Mang.
Local: Co-adviser: Roque, Reynaldo.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041062:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041062/00001

Material Information

Title: Determination of Test Methods for the Prediction of the Behavior of Mass Concrete
Physical Description: 1 online resource (277 p.)
Language: english
Creator: Ferraro, Christopher
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: calorimetry, concrete, heat, hydration, isothermal, mass, modulus, semi, strength
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Hydration at early ages results from chemical and physical processes that take place between Portland cement and water, and is an exothermic process. The resultant heat evolution and temperature rise for massive concrete placements can be so great that the temperature differentials between the internal concrete core and outer concrete stratum can cause cracking due to thermal gradients. Accurate prediction of temperature distribution and stresses in mass concrete is needed to determine if a given concrete mixture design may have problems in the field, so that adjustments to the design can be made prior to its use. This research examines calorimetric, strength, and physical testing methods in an effort to predict the thermal and physical behavior of mass concrete. Four groups of concrete mixture types containing different cementitious materials are examined. One group contains Portland cement, while the other three groups incorporate large replacements of supplementary cementitious materials: granulated blast furnace slag, fly ash, and a ternary blend (combining Portland cement, fly ash, and slag).
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Christopher Ferraro.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Tia, Mang.
Local: Co-adviser: Roque, Reynaldo.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041062:00001


This item has the following downloads:


Full Text

PAGE 1

1 DETERMINATION OF TEST METHODS FOR THE PREDICTION OF THE BEHAVIOR OF MASS CONCRETE By CHRISTOPHER C. FERRARO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIR EMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Christopher C. Ferraro

PAGE 3

3 To Jennifer L. Clark thanks for patience, support and Love

PAGE 4

4 ACKNOWLEDGMENTS I would like to expr ess my deepest gratitude to my advisor and committee chair, Dr. Mang Tia. Thank you for all of your support, kind words, advice and direction during this research process. After you inherited this project several years after its initiation, you guided me t hrough the complex process which involved the modification of this study to its finality. Y ou were always available to listen and offer advice w hether the subject matter was personal or professional, which I most certainly appreciated. I am also grateful for t he assistance offered by Dr. Reynaldo Roque throughout this process Your involvement in this research certainly prompted me to find the necessary focus that this study required. Further, your input with respect to the fundamentals of laboratory tes ting and the description of physical testing was a significant contribution to my growth as a researcher. I would like to express appreciation to Dr. H.R. Hamilton III for his assistance throughout this project. Your availability to provide technical expe rtise and input in this study was of great value to the process Additionally your willingness to offer professional and career advice as well as share your personal experiences with me were very welcome and helpful throughout this endeavor As the only person to serve on my masters and PhD committees, your presence offered continuity throughout the atypical circumstances surrounding this process. The advice and guidance offered by Dr. Peter Ifju of my committee is acknowledged and appreciated. I would l ike to express gratitude to Dr. Andrew J. Boyd, Dr. Joseph W. Tedesco, and Dr. Bjorn Birgisson. Your involvement in this research project was certainly beneficial and valuable. Acknowledgement is expressed to Mr. John Gadja and Dr. Anton Schindler both of who m provided invaluable input with respect to the technical issues pertaining to mass concrete.

PAGE 5

5 Dr. Jonathon L. Poole, your experience and knowledge of calorimetry, apparent activation energy and the general subject of mass concrete has been invaluable. Th ank you for providing technical advice, spreadsheet assi s tance and personal encouragement throughout this process. I certainly appreciate your willingness to consistently answer the phone despite your busy schedule Thanks to Dr. Christos Drakos for hi s assistance with the preparation of the final exam presentation, as well as the enormous support on the final exam day itself. The fact that you were willing to travel from Europe to assist me with the final exam was invaluable part of the success I achie ve d on that day. Your friendship and encouragement throughout this process will never be forgotten. I thank the following colleagues from the Florida Department of Transportation (FDOT) who provided assistance with this project. Thomas Malerk, Michael Ber gin, Mario Paredes, Christine McDonald, Richard Delorenzo, Joseph Fitzgerald, Toby Dillow, Craig Roberts, Steven Sauls, Fred Yon, Susan Blazo, Alfred Camps, Duane Robertson, Dr. H. Deford, Rahman Henderson, Teresa Risher and Sue Rose. Mostly, Id also like to thank Charles Ishee for dedicating a great deal of personal time and interest to this project. The direct assistance with experimental design, laboratory testing and general interest was extremely helpful, as well as your personal insight and critical assessment of the details of this research. Johann Kepler said I much prefer the sharpest criticism of a single intelligent man to the thoughtless approval of the masses. T hanks for contributing your continuous critical assessment of every aspect of this research entailed. Your knowledge of Portland cement concrete materials, and your friendship throughout this process was an invaluable pa rt of completing this research.

PAGE 6

6 I acknowledge the following colleagues from the University of Florida who assisted wit h this project. Tony Murphy, Charles Broward, Hubert Martin. Additionally, Id like to thank the following graduate students for assistance ; Adrian Lawrence, who was an excellent research partner. Thanks for remaining so pleasant and positive for what coul d otherwise have been a relatively unpleasant process. Additionally, Id like to thank the Boris Haranki, John Emery and who are fellow graduate students and good friends as well. Thanks for your assistance on the project itself and for personal encourag ement throughout this process. I also acknowledge the following graduate and undergraduate students for assistance with this project; Eiji Ochiai, Christopher Egan James Falls and Gustavo Morris. Id like to thank my friends and colleagues Marjorie Lynch and Thomas Howard who spent an enormous amount of personal time editing this document for syntax, clarity and content. Furthermore, I thank Marjorie for her mentorship and friendship throughout the years. Your, knowledge of the engineering industry, const ruction practices and common hatred of the Georgia Bulldogs and Florida State Seminoles football programs have always made conversations fruitful and meaningful. I thank Thomas for being a trusted friend whos availability for lunch at any hour (especially within the final stages of this process) good conversation, and common hated of the common hatred of the Georgia Bulldogs, Florida State Seminoles and Tennessee Volunteers football programs have always made conversations fruitful and meaningful. Id like to thank the following friends and colleagues who were a great support system throughout this process: Adam Smith, Richard Davis, Dr. Forrest Masters, Dr. David Prevatt, Tanya Nash, Thomas Currie, Gregory Doig, Dr. Daniel Algernon, Dr. Alvaro Guarin, Dr. Ulas Toros, James Hill, and Thomas Berke.

PAGE 7

7 I thank my family for all of their support and encouragement throughout this process. Thanks to my father Ronald Ferraro, my mother Victoria Ferraro, my brother Ronald Ferraro Jr. my sister Alicia Ferraro, my sist er -in -law Lucille Ferraro and my grandfather Charles Stokes. Anthony Brandt said other things may change us, but we start and end with family. Most importantly, I like to thank My Fianc Ms. Jennifer L. Clark for her unyielding patience, support, and love throughout this process. This research endeavor was a long, arduous, stressful process which required attention and focus that words cannot describe. Your selflessness as a partner and friend during this time of my personal, professional and academic ac complishment will never be forgotten Thank you Belle for everything. Mother Teresa said The greatest science in the world; in heaven and on earth; is love. Ive learned so much and yet have so much to learn. I pray our primary topic of study in the days to come is the science of love between you and I.

PAGE 8

8 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES .............................................................................................................................. 12 LIST OF FIGURES ............................................................................................................................ 15 ABSTRACT ........................................................................................................................................ 25 CHAPT ER 1 INTRODUCTION ....................................................................................................................... 26 Background .................................................................................................................................. 26 Definition of Mass Concrete ....................................................................................................... 26 Research Needs ........................................................................................................................... 28 Objectives of Research ............................................................................................................... 28 Hypothesis ................................................................................................................................... 29 Research Approach ..................................................................................................................... 29 Significance of Research ............................................................................................................ 30 Chemical Analysis of Cementitious Components ..................................................................... 30 Outline of Dissertation ................................................................................................................ 32 2 LITERATURE REVIEW ........................................................................................................... 34 Mass Concrete Specifications ..................................................................................................... 34 Tensile Stresses in Mass Concrete ............................................................................................. 39 Tensile Stresses Due to Thermal Gradients ....................................................................... 39 Tensile Stresses Due to Delayed Ettringite Formation ..................................................... 41 3 ISOTHERMAL CONDUCTION CALORIMETRY TESTING ............................................. 47 Introduction ................................................................................................................................. 47 Summary of Test Method ........................................................................................................... 48 Equipment and Procedure ........................................................................................................... 49 Mixing Procedure ........................................................................................................................ 49 External Mixing ................................................................................................................... 50 Internal Mixing .................................................................................................................... 50 Isothermal Calorimetry Testing Methodology .......................................................................... 51 Isothermal Calorimetry Testing Results .................................................................................... 52 Summary of Results .................................................................................................................... 55

PAGE 9

9 4 MATURITY AND EQUIVALENT AGE ................................................................................. 57 Maturity Concept ......................................................................................................................... 57 Equivalent Age Concept ............................................................................................................. 58 Arrhenius Equation and Activation Energy............................................................................... 59 Activation Energy Determination via Compressive Strength .................................................. 61 Apparent Activation Energy Determination as Per ASTM C 1074 ................................. 63 Activation Energy Determination Via Exponential Method ............................................. 65 Activation Energy Determination via Isothermal Calorimetry ................................................ 69 Comparison of Strength Development Parameters ........................................................... 74 Summary of Results Apparent Activation Energy ......................................................... 77 Summa ry of Findings Chapter 4 ................................................................................................ 79 5 SEMI -ADIABATIC CALORIMETRY TESTING .................................................................. 80 Introduction ................................................................................................................................. 80 Research Significance ................................................................................................................. 80 Equipment and Procedure ........................................................................................................... 81 Calibration Procedure .......................................................................................................... 81 Calculation of Adiabatic Temperature Rise ....................................................................... 83 Quadrel IQ drum test procedure .............................................................................. 83 University of Texas / Auburn University test procedure ........................................... 85 Semi -Adiabatic Calorimetry Testing Results ............................................................................ 85 Summary of Findings .................................................................................................................. 91 6 SURE CURE/ADIABATIC CALORIMETRY TESTING ...................................................... 92 Research Sign ificance ................................................................................................................. 92 Equipment .................................................................................................................................... 93 Test Procedure ............................................................................................................................. 94 Sure Cure Calorimetry Testing Results ..................................................................................... 96 Limitations of the Test Method .................................................................................................. 98 Summary of Findings .................................................................................................................. 98 7 COMPARISON OF CALORIMETRY TESTING METHODS ............................................ 100 Introduction ............................................................................................................................... 100 Research Significance ............................................................................................................... 100 Calorimetry Testing Results ..................................................................................................... 101 Comparison of Sure Cure adiabatic calorimeter and semi adiabatic calorimeter ......... 101 Comparison of semi adiabatic calorimeter and the isothermal calorimeter ................... 102 Comparison of cement and concrete ......................................................................... 103 Mortar testing ............................................................................................................. 108 Comparison of solution calorimetry to isothermal and semi adiabatic calorimetry ..... 112 Summary of Findings ................................................................................................................ 114 8 LARGE -SCALE BLOCK EXPERIMENT PH YSICAL TESTING ................................... 115

PAGE 10

10 Mixture Design .......................................................................................................................... 115 Concrete Specimen Creation and Curing ................................................................................ 116 Maturity and Equivalent Age ................................................................................................... 117 Maturity Calculation .......................................................................................................... 119 Equivalent age calculation ................................................................................................ 120 Physical Data and Results ......................................................................................................... 121 Compressive Strength Testing .......................................................................................... 121 Compressive strength vs. time ................................................................................... 121 Compressive strength vs. maturity ............................................................................ 122 Compressive strength vs. equivalent age .................................................................. 124 Splitting Tensile Strength Testing .................................................................................... 129 Splitting tensile s trength vs. time .............................................................................. 129 Tensile strength vs. maturity ..................................................................................... 130 Tensile strength vs. equivalent age ........................................................................... 130 Compressive Modulus of Elasticity .................................................................................. 134 Tangent modulus ........................................................................................................ 135 Secant modulus ........................................................................................................... 136 Chord modulus ........................................................................................................... 136 Modulus of elasticity methodology ........................................................................... 136 Compressive modulus of elasticity vs. time ............................................................. 138 Compressive modulus of elasticity vs. maturity ...................................................... 139 Compressive modulus of elasticity vs. equivalent age ............................................ 140 Tensile Modulus of Elasticity ........................................................................................... 144 Tensile modulus of elas ticity vs. time ....................................................................... 145 Comparison of tensile and compressive modulus of elasticity ............................... 146 Summary of Findings ................................................................................................................ 149 9 LARGE -SCALE BLOCK EXPERIMENT THERMAL TESTING .................................. 150 Coeffic ient of Thermal Expansion Testing ............................................................................. 150 Thermal Diffusivity Testing ..................................................................................................... 152 Specific Heat Capacity Testing ................................................................................................ 155 Summary of Findings ................................................................................................................ 157 10 RECOMMENDED TESTING PROGRAM FOR M ASS CONCRETE ............................... 158 Recommended Laboratory Testing Method for Measurement of Heat Generation ............. 158 Recommended Laboratory Testing Method for Measuring Maturity/Equivalent Age ........ 159 Recommended Laboratory Testing Methods for Strength and Modulus of Elasticity ......... 160 Compressive strength ........................................................................................................ 160 Compressive Modulus of Elasticity .................................................................................. 161 Tensile Strength ................................................................................................................. 161 Recommended Laboratory Testing Method for Measurement of Coefficient of Thermal Expansion ............................................................................................................................... 162 Recommended Physical Parameter for the Concrete Diffusivity ........................................... 162 Recommended Physical Paramete r for the Specific Heat Capacity ....................................... 163 Summary of Testing Program .................................................................................................. 163

PAGE 11

11 11 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK ............................. 164 Concl usions ............................................................................................................................... 164 Recommendations ..................................................................................................................... 164 APPENDIX A METHOD OF TESTING FOR MEASURING THE HEAT OF HYDRATION OF HYDRAULIC CEMENT USING ISOTHERMAL CONDUCTIVE CALORIMETRY .... 166 B RAW DATA MORTAR CUBES AND APPARENT ACTIVATION ENERGY CALCULATIONS .................................................................................................................... 172 C ISOTHERMAL CALORIMETRY RAW DATA AND ANALYSIS ................................... 185 Calorimetry Data of Cementitious Specimens ........................................................................ 185 Calorimetry Data of Mortar Specimens ................................................................................... 198 D SEMI ADIABATIC TEMPERATURE DATA ...................................................................... 207 E COMBINED CALORIMETRY DATA AND ANALYSIS .................................................. 211 F PHYSICAL DATA ................................................................................................................... 219 G CEMENTITIOUS MATERIAL DATA AND MILL CERTIFICATIONS .......................... 252 LIST OF REFERENCES ................................................................................................................. 267 BIOGRAPHICAL SKETCH ........................................................................................................... 277

PAGE 12

12 LIST OF TABLES Table page 2-1 FDOT bridge environmental database results ...................................................................... 35 2-2 Cementitous material content for concrete materials ........................................................... 35 2-3 Mass concrete requirements per state agency ...................................................................... 36 2-4 Typical cementitous content used for mass concrete ........................................................... 38 4-1 Mortar cube mixture designs ................................................................................................. 62 4-2 Strength development parameters for each concrete mixture per ASTM 1074 ................. 65 4-3 St rength development parameters for each concrete mixture per exponential method ..... 68 4-4 Strength development parameters for each concrete mixture per isothermal calorimetry / expoenential method ........................................................................................ 73 4-5 Strength development parameters for mixture 1 (100% Portland cement) per each method of analysis .................................................................................................................. 74 4-6 Summary of Ea values per testing and analysis method ...................................................... 77 4-7 Percentage differences of Ea per testing and analysis method ............................................ 78 5-1 Mixture designs for concrete specimens ............................................................................... 86 5-2 Delivery temperatures for concrete mixes ............................................................................ 87 7-1 Comparison of energy rise values for each calorime try method for each mix ................. 113 8-1 Mixture designs for large -scale block specimens .............................................................. 115 8-2 Summary of apparent activation energy pe r testing and analysis method ........................ 120 9-1 Specific heat for concrete material ...................................................................................... 156 A-1 Raw data sheet Mix 1 cubes 100% Portland cement ......................................................... 172 A-2 Raw data sheet Mix 1 cubes 100% Portland cement ......................................................... 172 A-3 Raw data sheet Mix 1 cubes 100% Portland cement ......................................................... 173 A-4 Raw data sheet Mix 2 cubes 50% Portland cement 50% slag ........................................ 175 A-5 Raw data sheet Mix 2 cubes 50% Portland cement 50% slag ........................................ 175

PAGE 13

13 A-6 Raw data sheet Mix 2 cubes 50% Portland cement 50% slag ........................................ 176 A-7 Raw data sheet Mix 3 cubes 65% Portland cement 35% fly ash ................................... 178 A-8 Raw data sheet Mix 3 cubes 65% Portland cement 35% fly ash ................................... 178 A-9 Raw data sheet Mix 3 cubes 65% portland cement 35% f ly ash ................................... 179 A10 raw data sheet Mix 4 cubes 50%30%20% blend (Cement slag fly ash) .................. 181 A11 Raw data sheet Mix 4 cubes 50%-3 0%20% blend (Cement slag fly ash) ................ 181 A12 Raw data sheet Mix 4 cubes 50%30%20% blend (Cement slag fly ash) ................ 182 C-1 M ixture components, respective apparent activation energy and datum temperatures ... 185 D-1 Mixture designs for concrete specimens ............................................................................. 207 E-1 Mixture designs for mortar specimens used in the semi adiabatic calorimeter ............... 211 F-1 Raw data compressive strength of cylinders Mix 1 100% Portland cement ................ 219 F-2 Raw data compressive strength of cylinders Mix 2 50% Portland cement 50% slag ........................................................................................................................................ 221 F-3 Raw data compressive strength of cylinders Mix 2 65% P ortland cement 35% fly ash .......................................................................................................................................... 223 F4. Raw data compressive strength of cylinders Mix 2 50% Portland cement 30% slag 20% fly ash ........................................................................................................................ 225 F-5 Raw data splitting tensile strength of cylinders Mix 1 100% Portland cement ............ 228 F-6 Raw data splitting tensile strength of cylinders Mix 2 50% Portland cement 50% slag ........................................................................................................................................ 230 F-7 Raw data splitting tensile strength of cylinders Mix 2 65% Portland cement 35% fly ash .................................................................................................................................... 232 F-8 Raw data splitting te nsile strength of cylinders Mix 4 50% Portland 30% slag 20% fly ash ........................................................................................................................... 234 F-9 Raw data compressive modulus of elasticity of cylinders Mix 1 100% Portland cement ................................................................................................................................... 237 F10 Raw data modulus of elasticity of cylinders Mix 2 50% Portland cement 50% slag ........................................................................................................................................ 239

PAGE 14

14 F11 Raw data modulus of elasticity of cylinders Mix 2 65% Portland cement 35% fly ash .......................................................................................................................................... 241 F12 Raw data modulus of elasticity of cylinders Mix 4 50% Portland cement 30% slag 20% fly ash .................................................................................................................. 243 F13 Raw data tensile modulus of elasticity of cylinders Mix 1 100% Portland cement ..... 246 F14 Raw data tensile modulus of elasticity of cylinders Mix 2 50% Portland ce ment 50% slag ................................................................................................................................ 247 F15 Raw data tensile modulus of elasticity of cylinders Mix 2 65% Portland cement 35% fly ash ........................................................................................................................... 248 F16 Raw data tensile modulus of elasticity of cylinders Mix 4 50% Portland 30% slag 20% fly ash .......................................................................................................................... 249 F17 Raw data coefficient of thermal expansion testing ............................................................ 251 G-1 Xray fluorescence data for Portland cement used in each mix ........................................ 254 G-2 Xray diffraction data for Portland cement used in each mix ........................................... 254

PAGE 15

15 LIST OF FIGURES Figure page 1-1 Behavior of mass concrete ..................................................................................................... 27 3-1 Isothermal conduction calorimeter test set up ...................................................................... 49 3-2 Mixing cell and ampoule used for internal mixing procedure ............................................ 50 3-3 Power vs. time for 0.50 w/c Portland cement at v arious temperatures ............................... 53 3-4 Energy vs. time for 0.50 w/c Portland cement at various temperatures ............................. 53 3-5 Fully hydrated specimen ( 23C) vs. partially hydrated specimen (60C) .......................... 54 3-6 Energy vs. time for 0.40 w/c Portland cement at various temperatures ............................. 55 4-1 Nurse Saul maturity function (Carino, 2004) ....................................................................... 58 4-2 Compressive strength of mortar cubes vs. time for each temperature ................................ 64 4-3 Act ivation energy calculation for hyperbolic method ......................................................... 65 4-4 Compressive strength of mortar cubes vs. time for each temperature ................................ 68 4-5 Activation energy calculation for exponential method ........................................................ 69 4-6 Isothermal calorimetry data vs. time for four temperatures ............................................. 72 4-7 Apparent activation energy calculation for exponential method ........................................ 73 4-8 Residual error vs. normalized time for hyperbolic curve fitting method (compressive strength) .................................................................................................................................. 75 4-9 Residual error vs. normalized time for exponential curve fitting method (compressive strength) .................................................................................................................................. 76 410 Residual error vs. normalized time for exponential curve fitting method (isothermal calorimetry) ............................................................................................................................ 76 5-1 Heat flux sensor ...................................................................................................................... 82 5-2 Calculated losses for the water calibration test .................................................................... 83 5-3 Adiabatic temperature rise for Mix 1 .................................................................................... 87 5-4 Adiabatic temperature rise curves for Mix 1, delivered mix and lab mix .......................... 89

PAGE 16

16 5-5 Adiabatic temperature rise for Mix 2, delivered mix and lab mix ...................................... 89 5-6 Adiabatic temperature rise for each mix ............................................................................... 90 6-1 Schematic of Sure Cure system with hydration chamber .................................................... 93 6-2 Photograph of the Sure Cure system with the computer, hydration chamber and controller box .......................................................................................................................... 94 6-3 Sure Cure and semi adiabatic temperature vs. time ............................................................. 96 6-4 Sure Cure and semi adiabatic temperature vs. time ............................................................. 97 6-5 Sure Cure temperature rise for large scale block experiments ............................................ 97 7-1 Comparison of time temperature histories between Sure Cure system and the IQ drum for Mix 1 ..................................................................................................................... 101 7-2 Comparison of time temperature historie s between Sure Cure system and the IQ drum for Mix 2 ..................................................................................................................... 102 7-3 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 1 ..................................................................... 104 7-4 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 2 ..................................................................... 105 7-5 Mortar cube compressive strength vs. time for Mix 2 ....................................................... 106 7-6 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 3 ..................................................................... 107 7-7 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 4 ..................................................................... 107 7-8 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and the IQ drum for mortar Mix 1 .......................................................................... 109 7-9 Comparison of equivalent age energy hi stories between semi adiabatic calorimetry system and the IQ drum for mortar Mix 3 .......................................................................... 109 710 Cross section of isothermal calorimetry device ................................................................. 110 711 Cross section of semi adiabatic device ............................................................................... 112 8-1 Large scale block testing configuration .............................................................................. 116 8-2 Thermocouple location for Sure Cure system .................................................................... 117

PAGE 17

17 8-3 Time/temperature history of 4x8 cylinder specimens .................................................... 118 8-4 Time/temperature history bloc k Sure Cure cylinder specimens .................................... 119 8-5 Compressive strength vs. time for each mix ....................................................................... 121 8-6 Compressive strength vs. time for Mi x 1 100% Portland cement .................................. 122 8-7 Compressive strength vs. maturity for Mix 1 100% Portland cement ........................... 123 8-8 Compressive streng th vs. equivalent age for Mix 1 100% Portland cement ................. 124 8-9 Sure Cure specimen mold .................................................................................................... 125 810 Compressive strength vs. equival ent age for Mix 2 50% Portland cement 50% slag ........................................................................................................................................ 127 811 Compressive strength vs. equivalent age for Mix 3 65% Portland cement 35% fly ash .......................................................................................................................................... 128 812 Compressive strength vs. equivalent age for Mix 4 50% Portland cement 30% slag 20% fly ash .......................................................................................................................... 128 813 Splitting tensile strength vs. time for each mix .................................................................. 129 814 Splitting tensile strength vs. maturity for Mix 1 100% Portland cement ....................... 130 815 Splitting tensile strength vs. equivalent age for Mix 1 100% Portland cement ............. 131 816 Splitting tensile strength vs. equivalent age for Mix 2 50% Portland cement 50% slag ........................................................................................................................................ 132 817 Splitting tensile strength vs. equivalent age for Mix 3 65% Portland cement 35% fly ash .......................................................................................................................................... 133 818 Splitting tensile strength vs. equivalent age for Mix 4 50% Portland cement 30% sl ag 20% fly ash .................................................................................................................. 133 819 Typical stress -strain diagram for concrete, showing the different elastic moduli (Mindess 2003) ..................................................................................................................... 135 8-20 Stress -strain diagram for concrete specimen ...................................................................... 137 821 Compressive modulus of elasticity vs. time for each mix ................................................. 139 822 Compress ive modulus of elasticity vs. maturity for Mix 1 100% Portland cement ....... 140 823 Compressive modulus of elasticity vs. equivalent age for Mix 1 100% Portland cement ................................................................................................................................... 141

PAGE 18

18 824 Compressive modulus of elasticity vs. equivalent age for Mix 2 50% Portland cement 50% slag ................................................................................................................ 142 825 Compressive modulus of elasticity vs. equivalent age equivalent age for Mix 3 65% Portland cement 35% fly ash............................................................................................. 143 826 Compressive modulus of elasticity vs. equivalent age for Mix 4 50% Portland cement 30% slag 20% fly ash .......................................................................................... 143 827 Photo of tensile (flexural) modulus of elasticity test configuration .................................. 144 828 Tensile modulus of elasticity vs. time for each mix .......................................................... 145 829 Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 1 ........................................................................................................................ 147 830 Ten sile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 2 ........................................................................................................................ 148 831 Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age fo r Mix 3 ........................................................................................................................ 148 832 Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 4 ........................................................................................................................ 149 9-1 Coeffi cient of thermal expansion vs. time for each mix .................................................... 151 9-2 Thermal diffusivity vs. time for each mix .......................................................................... 153 A-1 A schematic drawing o f a heat conduction calorimeter ..................................................... 167 A-2 Steady state calibration plot ................................................................................................. 168 A-3 Cutaway of one of the 8 calorimetric channels show ing the twin configuration ............. 171 B-1 Compressive strength vs. time hyperbolic model for Mix 1 (ASTM C 1074) .............. 173 B-2 Compressive strength vs. time exponential model for Mix 1 (Constant ................... 174 B-3 Compressive strength vs. time exponential model for Mix 1 (constant u .............. 174 B-4 Compressive strength vs. time hyperbolic model for Mix 2 (ASTM C 1074) .............. 176 B-5 Compressive strength vs. time exponential model for Mix 2 (Constant ................... 177 B-6 Compressive strength vs. time exponential model for Mix 2 (constant u .............. 177 B-7 Compressive strength vs. time hyperbolic mode l Mix 3 (ASTM C 1074) ................. 179

PAGE 19

19 B-8 Compressive strength vs. time exponential model for Mix 3 (Constant ................... 180 B-9 Compressive str ength vs. time exponential model for Mix 3 (constant u .............. 180 B10 Compressive strength vs. time hyperbolic model Mix 4 (ASTM C 1074) ................. 182 B11 Compressive strength vs. time exponential model for Mix 4 (Constant ................... 183 B12 Compressive strength vs. time exponential model for Mix 4 (constant u .............. 183 B13 Summary of time of set data ................................................................................................ 184 C-1 Power vs. time for Mix 1 ..................................................................................................... 186 C-2 Energy vs. time for Mix 1 .................................................................................................... 186 C-3 Power vs. maturity for Mix 1 .............................................................................................. 187 C-4 Energy vs. maturity for Mix 1 ............................................................................................. 187 C-5 Power vs. equivalent age for Mix 1 .................................................................................... 1 88 C-6 Energy vs. equivalent age for Mix 1 ................................................................................... 188 C-7 Power vs. time for Mix 2 ..................................................................................................... 189 C-8 Energy vs. time for Mix 2 .................................................................................................... 189 C-9 Power vs. maturity for Mix 2 .............................................................................................. 190 C10 Energy vs. maturity for Mix 2 ............................................................................................. 190 C11 Power vs. equivalent age for Mix 2 .................................................................................... 191 C12 Energy vs. equivalent age for Mix 2 ................................................................................... 191 C13 Power vs. time for Mix 3 ..................................................................................................... 192 C14 Energy vs. time for Mix 3 .................................................................................................... 192 C15 Power vs. maturity for Mix 3 .............................................................................................. 193 C16 Energy vs. maturity for Mix 3 ............................................................................................. 193 C17 Power vs. equivalent age for Mix 3 .................................................................................... 194 C18 Energy vs. equivalent age for Mix 4 ................................................................................... 194 C19 Power vs. time for Mix 4 ..................................................................................................... 195

PAGE 20

20 C20 Power vs. energy for Mix 4 ................................................................................................. 195 C21 Power vs. maturity for Mix 3 .............................................................................................. 196 C22 Energy vs. maturity for Mix 4 ............................................................................................. 196 C23 Power vs. equivalent age for Mix 4 .................................................................................... 197 C24 Energy vs. equivalent age for Mix 4 ................................................................................... 197 C25 Power vs. time for Mortar Mix 1 ......................................................................................... 198 C26 Energy vs. time for Mortar Mix 1 ....................................................................................... 199 C27 Power vs. maturity for Mortar Mix 1 .................................................................................. 199 C28 Power vs. maturity for Mortar Mix 1 .................................................................................. 200 C29 Power vs. e quivalent age for Mortar Mix 1 ........................................................................ 200 C30 Energy vs. equivalent age for Mortar Mix 1 ...................................................................... 201 C31 Power vs. time for Mortar Mix 3 ......................................................................................... 201 C32 Energy vs. time for Mortar Mix 3 ....................................................................................... 202 C33 Power vs. maturity for Mortar Mix 3 .................................................................................. 202 C34 Energy vs. maturity for Mortar Mix 3 ................................................................................ 203 C35 Power vs. equivalent age for Mortar Mix 3 ........................................................................ 203 C36 Ene rgy vs. equivalent age for Mortar Mix 3 ...................................................................... 204 C37 Isothermal calorimetry data and plots used for the calculation of activation energy for Mix 1 ............................................................................................................................... 204 C38 Isothermal calorimetry data and plots used for the calculation of activation energy for Mix 2 ............................................................................................................................... 205 C39 Isothermal calorimetry data and plots used for the calculation of activation energy for Mix 3 ............................................................................................................................... 205 C40 Isothermal calorimetry data and plots used for the calculation of activation energy for Mix 4 ............................................................................................................................... 206 D-1 Typical water calibration ..................................................................................................... 207 D-2 Semi adiabatic temperature measurements for mix 1 (laboratory mix) ........................... 208

PAGE 21

21 D-3 Semi ad iabatic temperature measurements for concrete mix 1 (lab and delivered) mixes ..................................................................................................................................... 208 D-4 Semi adiabatic temperature measurements for concrete mix 2 (lab and delivered) mixes ..................................................................................................................................... 209 D-5 Semi adiabatic temperature measurements for mix 3 (delivered) .................................... 209 D-6 Semi adiabatic temperature measurements for mix 4 (delivered) .................................... 210 D-7 Semi adiabatic temperature measurements for each mix .................................................. 210 E-1 Comparison of the time temperature histories between the Sure Cure s ystem and the IQ drum for Mix 1 ................................................................................................................ 212 E-2 Comparison of the time temperature histories between the Sure Cure system and the IQ drum for Mix 2 ................................................................................................................ 212 E-3 Comparison of the time temperature histories between the Sure Cure system and the IQ drum for Mix 3 ................................................................................................................ 213 E-4 Comparison of the time temperature histories between the Sure Cu re system and the IQ drum for Mix 4 ................................................................................................................ 213 E-5 Time -temperature histories obtained by the Sure Cure for each mix ............................... 214 E-6 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 1. .................................................................... 214 E-7 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 2. .................................................................... 215 E-8 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for M ix 3. .................................................................... 215 E-9 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 4. .................................................................... 216 E10 Comparison of equivalent age energy histories between semi adiabatic calorimetry system and the IQ drum for mortar Mix 1. ......................................................................... 216 E11 Comparison of equivalent age energ y histories between semiadiabatic calorimetry system and the IQ drum for mortar Mix 3. ......................................................................... 217 E12 Comparison of semi adiabatic test and three different Sure Cure system offset temperatures for a trial mix. ................................................................................................ 217 E13 Comparison of semi adiabatic test and four different Sure Cure system offset temperatures for a trial mix. ................................................................................................ 218

PAGE 22

22 F-1 Compressive strength vs. time for Mix 1 100% Portland cement .................................. 220 F-2 Compressive strength vs. maturity for Mix 1 100% Portland cement ........................... 220 F-3 Compressive strength vs. equivalent age for Mix 1 100% Portland cement ................. 221 F-4 Compressive strength vs. time for Mix 2 50% Portland cement 50% slag ................. 222 F-5 Compressive strength vs. maturity for Mix 2 50% Portland cement 50% slag .......... 222 F-6 Compressive strength vs. equivalent age for Mi x 2 50% Portland 50% slag ............. 223 F-7 Compressive strength vs. time for Mix 3 65% Portland cement 35% fly ash ............ 224 F-8 Compre ssive strength vs. maturity for Mix 3 65% Portland cement 35% fly ash ..... 224 F-9 Compressive strength vs. equivalent age for Mix 3 65% Portland cement 35% fly ash .......................................................................................................................................... 225 F10 Compressive strength vs. time for Mix 4 50% Portland 30% slag 20% fly ash ........ 226 F11 Compressive strength vs. maturity for Mix 4 50% Portl and 30% slag 20% fly ash 226 F12 Compressive strength vs. equivalent age for Mix 4 50% Portland 30% slag 20% fly ash .................................................................................................................................... 227 F13 Compressive strength vs. age summary for each mix ........................................................ 227 F14 Compressive strength vs. equivalent age summary for each mix ..................................... 228 F15 Tensile strength vs. time for Mix 1 100% Portland cement ............................................ 229 F16 Tensile strength vs. maturity for Mix 1 100% Portland cement ..................................... 229 F17 Tensile strength vs. equivalent age for Mix 1 100% Portland cement ........................... 230 F18 Tensile strength vs. time for Mix 2 50% Portland cement 50% slag .......................... 231 F19 Tensile strength vs. maturity for Mix 2 50% Portland cement 50% slag ................... 231 F20 Tensile strength vs. equivalent age for Mix 2 50% Portland c ement 50% slag ......... 232 F21 Tensile strength vs. time for Mix 3 65% Portland cement 35% fly ash ......................... 233 F22 Tensile strength vs. ma turity for Mix 3 65% Portland cement 35% fly ash .................. 233 F23 Tensile strength vs. equivalent age for Mix 3 65% Portland cement 35% fly ash ........ 234 F24 Tensile strength vs. time for Mix 4 50% Portland 30% slag 20% fly ash ................ 235

PAGE 23

23 F25 Tensile strength vs. maturity for Mix 4 50% Portland 30% slag 20% fly ash ......... 235 F26 Splitting tensile strength vs. equivalent age for Mix 4 50% Portland 30% slag 20% fly ash ........................................................................................................................... 236 F27 Compressive strength vs. age summary for each mix ........................................................ 236 F28 Tensile strength vs. equivalent age summary for each mix ............................................... 237 F29 Compressive modulus of e lasticity vs. time for Mix 1 100% Portland cement .............. 238 F30 Compressive modulus of elasticity vs. maturity for Mix 1 100% Portland cement ....... 238 F31 Compressive modulus of elasticity vs. equivalent age for Mix 1 100% Portland cement ................................................................................................................................... 239 F32 Compressive modulus of elasticity vs. age for Mix 2 50% Portland cement 50% slag ........................................................................................................................................ 240 F33 Compressive modulus of maturity vs. age for Mix 2 50% Portland cement 50% slag 240 F34 Compressive modulus of e lasticity vs. equivalent age for Mix 2 50% Portland cement 50% slag ................................................................................................................ 241 F35 Compressive modulus of elasticity vs. age for Mix 3 65% Portland cement 35% fly ash .......................................................................................................................................... 242 F36 Compressive modulus of maturity vs. age for Mix 3 65% Portland cement 35% fly ash .......................................................................................................................................... 242 F37 Compressive modulus of elasticity vs. equivalent age equivalent age for Mix 3 65% Portland cement 35% fly ash ............................................................................................... 243 F38 Compressive modulus of elasticity vs. age for Mix 4 50% Portland cement 30% slag 20% fly ash ................................................................................................................. 244 F39 Compressive modulus of elasticity vs. maturity for Mix 4 50% Portland cement 30% slag 20% fly ash ........................................................................................................ 244 F40 Compressive modulus of elasti city vs. equivalent age for Mix 4 50% Portland cement 30% slag 20% fly ash .......................................................................................... 245 F41 Compressive modulus of elasticity vs. age for each mix................................................... 245 F42 Compressive modulus of elasticity vs. age for each mix................................................... 246 F43 Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 1 100% Portland cement ............................................................................... 247

PAGE 24

24 F44 Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 2 50% Portland cement 50% slag................................................................. 248 F45 Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 3 65% Portland cement 35% fly ash ............................................................ 249 F46 Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 4 50% Portland 30% slag 20% fly ash .................................................... 250 F47 Tensile modulus of elasticity vs. time for each mix .......................................................... 250 F48 Coefficient of thermal expansion vs. time for each mix .................................................... 251 G-1 Solution calorimetry data for Mix 1 100% Portland cement .......................................... 252 G-2 Solution calorimetry data for Mix 2 50% Portland cement 50% slag ........................ 252 G-3 Solution calorimetry data for Mix 3 65% Portland c ement 35% fly ash.................... 253 G-4 Solution calorimetry data for Mix 3 50% Portland cement 30% slag 20% fly ash .......................................................................................................................................... 253 G-5 Laborator y testing data of Portland cement used for Mix 1 .............................................. 255 G-6 Laboratory testing data of Portland cement used for Mix 1 .............................................. 256 G-7 Labor atory testing data of Portland cement used for Mix 2 .............................................. 257 G-8 Laboratory testing data of Portland cement used for Mix 2 .............................................. 258 G-9 L aboratory testing data of granulated blast furnace slag used for Mix 2 ......................... 259 G10 Laboratory testing data of granulated blast furnace slag used for Mix 2 ......................... 260 G11 Laboratory testing data of Portland cement used for Mix 3 .............................................. 261 G12 Laboratory testing data of fly ash used for Mix 3 .............................................................. 262 G13 Laboratory testing data of fly ash used for Mix 3 .............................................................. 263 G14 Mill certification data for Portland cement used in each mix ........................................... 264 G15 Mill certification data for granulated blast furnace slag cement used in each mix .......... 265 G16 Mill certification data for fly ash cement used in each mix .............................................. 266

PAGE 25

25 Abstract of Dissertation Presented to the Graduate School of the University of Florid a in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DETERMINATION OF TEST ME THODS FOR THE PREDICTION OF THE BEHAVIOR OF MASS CONCRETE By Christopher C. Ferraro December 2009 Chair: Name Mang Tia Co chair : Reynaldo Roque Major: Civil Engineering H ydration at early ages results from chemical and physical processes that take place between Portland cement and water and is an exothermic process. The resultant heat evolution and temperature rise for massive concrete placements can be so great that the temperature differentials between the internal concrete core and outer concrete stratum can cause cracking due to thermal gradients. Accurate prediction of temperature distribution and stresses in mass concrete is needed to determine if a given concrete mixture design may have problems in the field so that adjustments to the design ca n be made prior to its use. This research examines calorimetric, strength and physical testing methods in an effort to predict the thermal and physical behavior of mass concrete. Four groups of concrete mixture types containing different cementitious mat erials are examined. One group contains Portland cement while the other three groups incorporate large replacements of supplementary cementitious materials : granulated blast furnace slag, fly ash, and a ternary blend ( combining Portland cement fly ash a nd slag).

PAGE 26

26 CHAPTER 1 INTRODUCTION Background The setting and hardening of traditional concrete are the result of chemical and physical processes that occur when Portland cement and water are combined. This chemical reaction (hydration) is an exothermic process which results in the evolution of heat, and the rise of concrete temperature. For mass concrete elements, the heat evolution and resultant temperature rise can be so great that the temperature differentials between the internal concrete core and out er concrete stratum result in concrete cracking due to thermal gradients. Accurate modeling for the prediction of temperature distribution and stresses in mass concrete is needed to determine if a given concrete mixture design may have problems in real wor ld applications so that adjustments to the mix design can be made prior to its adopted use. Definition of Mass Concrete Concrete cast in mass structures can develop excessive internal tensile stresses capable of inducing significant cracking damage and possible failure. Such cracking can lead to reduced structural capacity and certainly degrades concrete durability by permitting the ingress of deleterious ionic species such as chlorides and sulfates. The American Concrete Institute (ACI) currently defines mass concrete as any volume of concrete with dimensions large enough to require that measures be taken to cope with generation of heat from hydration from cement and attendant volume change to minimize cracking (ACI 207.1 R96). However this definition is a bit too vague for specification of concrete. State and local agencies typically define mass concrete within their construction specification documents. Specification requirements vary with the specifying organization and with local conditions.

PAGE 27

27 Fig ure 11. Behavior of mass concrete The primary reason mass concrete cracks is due to the heat evolution and resultant temperature differentials between the internal concrete core and outer concrete stratum. Figure 1 1 provides an illustration of evolution of cracking in concrete masses (Bamforth 1984). When concrete is placed, i t is a plastic material, uniform in temperature and has essentially no strength At early ages, the inner core of concrete will begin to develop heat and experience a resultant tem perature rise as illustrated in Figure 1 1a. However, the outer stratum will tend to experience localized cooling (with the absence of adequate insulation) due to the dissipation of heat to the colder environment. However, the inner core of the mass concrete will be insulated and the heat will not be dissipated causing a localized temperature rise as illustrated I figure 1 1b. The hotter

PAGE 28

28 concrete in the inner core experience s localized expansion shown by figure 1 1c. The resultant cracking will be tensile i n nature, typically initiating at the surface of the concrete and propagating inward, shown by figure 1 1d. Research Needs For the modeling and prediction of temperature distribution in mass concrete structures, knowledge of the following properties of c oncrete is needed: Heat generation rate Specific heat capacity Thermal diffusivity Coefficient of thermal expansion For the prediction of stresses, strains and the resistance to cracking of mass concrete structures, the following properties at e arly ages must also be known: Compressive strength Tensile and flexural strength Compressive modulus of elasticity Tensile modulus of elasticity Complications and uncertainties in the determination of these properties arise since these properties are functions of t he hydration state of the cementitious materials and are therefore dependent upon the curing history of the concrete. The industry needs a reliable and effective testing regimen and analysis method so that the results from laboratory tests can be used in t hermal and stress analysis for the prediction of the behavior of mass concrete. Objectives of Research An accurate determination of physical and thermal properties of early age concrete (as functions of the maturity/hydration of concrete) is necessary in order to predict the physical behavior a resulting from thermal distribution and corresponding stresses within mass concrete. The current models available isolate specific properties of mass concrete in effort to predict

PAGE 29

29 behavior. However, there are many p arameters which influence mass concrete behavior and these properties are interrelated and must be considered a system. The main objective of this study is to develop a test regimen and analysis methodology to provide these necessary input parameters. The specific objectives of this research are: Identify the appropriate test method for the prediction and measurement of heat of hydr ation in cementitious materials Analyze the results from heat of hydration testing which can be utilized for modeling and predicting temperature development. Establish quantifiable relationships between temperature history and physical properties of concrete materials. Develop a physical testing regimen which efficiently quantifies and characterizes the characteristics which are related to the thermal cracking of mass concrete. Hypothesis The performance of mass concrete in its resistance to cracking is governed by a set of critical field conditions which include heat of hydration of cementitious components resulting in tempe rature rise thermal characteristics, modulus of elasticity, and relative strength. Accurate prediction of the behavior of a mass concrete structure depends on the accuracy of the documentation of the thermal and mechanical properties used. In the past, th ese properties have typically been determined with respect to time. Earlyage thermal and mechanical properties of cementitious materials are functions of the maturity/hydration history of these materials and must be properly evaluated through laboratory t esting and data analysis. Research Approach Separate physical and thermal testing programs were developed to investigate state-of the art testing methods and to determine those methods which are suitable as input parameters for analysis of the behavior of mass concrete.

PAGE 30

30 The thermal testing regimen consisted of the following: Isothermal conduction calorimetry Semi adiabatic calorimetry Sure Cure /adiabatic calorimetry Heat of solution calorimetry (ASTM C186) Thermal diffusivity (CRD C 36 73) Specific heat capacity The physical testing regimen involved the following tests: Compressive strength (ASTM C39/109) Compressive modulus of elasticity (ASTM C469) Splitting tensile strength (ASTM C496) Flexural Strength (ASTM C 78) Tensile Modulus of elasticity Coeffi cient of thermal expansion (AASHTO TP60) The program included testing on relatively small cement and concrete specimens L arge scale block specimens were created in an effort to validate the measure the temperature and strain of large scale specimens created in the lab. A portion of the project involved the creation of a finite element model (Lawrence, 2009), and accordingly the results obtained from the physical testing were used to validate the model upon its completion. Significance of Research Current ly, there are no reliable and effective methods for determining the thermal and p hysical/strength properties that predict thermal and physical behavior of mass concrete at early ages. Therefore, development of a testing regimen and analysis technique which can produce the necessary thermal and physical properties of concrete used as model input parameters for mass concrete would constitute a substantial contribution to science and the construction industry. Chemical Analysis of Cementitious Components Rese archers often use cement chemistry to determine activation energy and total heat available in a cementitious constituent as provided by Xray fluorescence ( XRF) or X-ray

PAGE 31

31 diffraction ( XRD ) analysis (Schindler, 2004; Schindler and Folliard, 2005). The origin al equations used for the conversion of chemical oxides to the cement phases were developed by Bogue in the 1940s (Bogue, 1947). However, it has been recently determined that the chemical data obtained via XRF have relative bias and potential errors (Stut zman, 2007). Accordingly, chemical analysis for the prediction of heat of hydration may be error prone. The American Society of Testing and Materials ( ASTM ) and the American Association of State and Highway Officials ( AASHTO) recently harmonized their spe cifications for Portland cement in an effort to create uniformity in the standards. As a result, ASTM C150 and AASHTO M85 now have the same requirements regarding Portland cement. One of the major changes in the standard is that both standards no longer re quire a limit on C3S for type II (considered low heat) cement (Melander, 2007). Prior to the change, ASTM C 15004 table 2 limited the C3S content to be 58%. Currently, both standards require the type II cement to meet the following requirement (ASTM, 2008d): %100 75.43 3 ACSC (1-1) Typically, the components in Portland cement necessary for the calculation of equation 1 1 are obtained via X ray florescence (XRF) as per ASTM C 114, the standard test method for the chemical analysis of Portland cemen t. Note 4 of ASTM C15007 states The limit on the sum of C3S +4.75 C3A provides control on the heat of hydration of cement and is consistent with a test method C 186 7day heat of hydration limit of 335KJ/kg (80cal/g) (ASTM, 2008d, p2). However, the re quired quality assurance and quality control of cements used on FDOT projects, has revealed that several type I/II cements which adhere to equation 1 1 per XRF analysis were determined to have heat of hydration values at 7 days per ASTM C 186 above 335KJ/kg. Thus,

PAGE 32

32 the concept of limiting the heat of hydration of cement via the use of chemical analysis is not appropriate. Considering the documented inconsistencies in the chemical analysis of Portland cement components discussed in published documents t he u se of chemical analysis for the prediction of the heat of hydration of cementitious materials and Portland cement concrete does not yield conservative results. T herefore development of a finite element model which utilizes physical testing (calorimetry) a s opposed to chemical analysis for the determination of heat generation and temperature rise of concrete is proposed. Currently, no known analysis model predicts the heat rise in massive concrete structures and also incorporates the results provided by is othermal conduction calorimetry for input parameters. H owever, a model documented by Poole and Riding (Poole et. al, 2007 and Riding et. al, 2007) utilizes semi adiabatic calorimetry for the modeling of mass concrete. Both the semi adiabatic calorimetry t est and the isothermal calorimetry test were investigated in this study. Outline of Dissertation Chapter 2 is a review of the literature citing current specifications and sources for stresses which result in cracking due to temperature rise in massive co ncrete structures. Chapter 3 addresses the details of the isothermal cement calorimetry test and its potential uses for mass concrete. Chapter 4 discusses maturity and equivalent age determination of cementitious and concrete materials and the importance of determining their appropriate relationships. Chapters 5 and 6 discuss the semi adiabatic and adiabatic calorimetry, respectively. The chapters within this document discuss the thermal testing techniques that were studied in this project with the except ion of thermal diffusivity and specific heat capacity testing regimens. Q uantification of

PAGE 33

33 thermal diffusivity and specific heat capacity of concrete composed of mixture components typical for the state of Florida are reported by Ochiai (2009). Chapter 7 addresses the maturity and equivalent age comparisons between isothermal calorimetry and semi adiabatic and adiabatic calorimetry. Chapter 8 examines the results obtained from the physical testing regimen performed on the large block specimens. Chapter 9 d iscusses the results of the large block experiment testing. Recommendations for a testing program for the prediction of mass concrete behavior are detailed in Chapter 10. Chapter 11 provides discussion, conclusions, and recommendations for future research efforts.

PAGE 34

34 CHAPTER 2 LITERATURE REVIEW Mass Concrete Specifications More recently, ACI Specifications for Structural Concrete (ACI 30105) has made provisions pertaining to mass concrete within the Optional Requirements Checklist, which identifies actions required by the Architect/ Engineer to include the following; "Designate portions of the structure to be treated as either plain mass concrete or reinforced mass concrete. Whether or not concrete should be designated as mass concrete depends on many factor s, including weather conditions, the volume -surface ratio, rate of hydration, degree of restraint to volume change, temperature and mass of surrounding materials, and functional and aesthetic effect of cracking. In general, heat generation should be considered when the minimum cross sectional dimension approaches or exceeds 2 1/2 ft [760 mm] or when cement contents above 600 lb/yd [356 kg/m ] are used. The requirements for each project however, should be evaluated on their own merits." (ACI 301-R05) The ACI 301 document defines mass concrete by both a minimum dimension and a minimum cementitious content. However, to consider any concrete with a total cementitious content above 600 lb/yd3 to be mass concrete is not a practical requirement for concrete con struction in the state of Florida. Table 2.1 lists the structural classification s of concrete as per Section 346 of the FDOT Road and Bridge Construction. Per Table 2.1, any structural concrete with requirements above a Class II bridge deck is required to have a total cementitious content above 608 lb/yd3. The FDOT Structures Manual states all structural elements exposed to a moderately or severely aggressive environment must be made of Class IV, V or VI concrete. A general survey of the FDOTs bridge envi ronmental database was used to determine the approximate percentage of bridge structures which are located in moderately or severely

PAGE 35

35 aggressive environments in the State of Florida. Table 2.1 provides the number of bridges in Florida and the classification of each. A survey of the FDOTs Bridge Environmental Database was used to determine the approximate percentage of bridge structures within the State of Florida which are located in moderately or severely aggressive environments. Table 2.1 provides the total number and environmental classification of each bridge within the State of Florida. Table 2 1. FDOT b ridge e nvironmental d atabase r esults Total number of bridges in Florida* 11734 Number of b ridges in the FDOT Bridge Environmental Database 6800 Numb er of bridges in severely or mode rately aggressive environments 3262 Percentage of bridges in severely or moderately aggressive environments 48% Per National Bridge Inventory Records According to the FDOTs Bridge Environmental Database, the majority of bridges in the State of Florida are considered to be located within a moderately or severely aggressive environment. Therefore, if the construction industry within the state of Florida adopted the Optional Requirements Checklist from ACI 301 R05, then a large percentage of the structural concrete produced in the state would be considered mass concrete. Table 2 -2 Cementitous m aterial content for c oncrete m aterials Class of Concrete Minimum Total Cementitious Materials Content (lb/yd3) Maximum Water Ce mentitious Materials Ratio (lb/lb) I (Pavement) 508 0.50 I (Special) 508 0.50 II 564 0.49 II Bridge Deck 611 0.44 III 611 0.44 III (Seal) 611 0.52 IV 658 0.41 IV (Drilled Shaft) 658 0.41 V (Special) 752 0.37 V 752 0.37 VI 752 0.37

PAGE 36

36 Many state DOTs define mass concrete with a minimum dimension or in a volumetric manner similar to ACI 207. The Florida Department of Transportation Structures Manual defines mass concrete to be the following: 1. All Bridge components Except Drilled Shafts: When the minimum dimension of the concrete exceeds 36 -inches and the ratio of volume of concrete to the surface area is greater than 12 inches, provide for mass concrete. (The surface area for this ratio includes the summation of all the surface areas of the concret e component being considered, including the full underside (bottom) surface of footings, caps, construction joints, etc.) Note volume and surface area calculations in units of feet. 2. Drilled Shafts: All drilled shafts with diameters greater than 72 inches shall be designated as mass concrete and a Technical Special Provision may be required. (FDOT, 2009) Seventeen out of the fifty State DOTs have requirements incorporated into the definition and the conditions of mass concret e structural elements. Tab le 2.3 provides a summary of the requirements for mass concrete regarding minimum dimension for the classification of mass concrete elements, specifications for maximum placing temperature, maximum allowable curing temperature and maximum allowable tempera ture differential (Chini et.al, 2003). Table 2 -3 Mass c oncrete requirements p er s tate a gency State Minimum Dimension (ft.) Maximum Placement Temperature (F) Maximum Curing Temperature (F) Maximum Temperature Differential (F) Arkansas 75 36 California 6.5 160 Specified by thermal control plan provided by contractor Colorado 5 165 45 Delaware Determined on project basis 160 First 48 hrs= 40F 27 days= 50F 8 14 days= 60 F Florida 3 (structural) / 6(drill shaft) 185 35 Georgia 50 Illinois 160 35 Indiana 35

PAGE 37

37 Table 2 3. Continued State Minimum Dimension (ft.) Maximum Placement Temperature (F) Maximum Curing Temperature (F) Maximum Temperature Differential (F) Iowa 3.9 65 160 35 Kentucky 160 35 Min nesota 160 35 Nebraska 176 27 North Carolina 75 36 North Dakota 160 50 South Carolina 80 35 Texas 75 35 Virginia 5 170F (slag) 160F (fly ash) 35 West Virginia 4 160 35 As per T able 2 -3 the FDOT does not have requirements on the maximum placement temperature nor the maximum allowable temperature in mass concrete. The FDOT requires that a temperature control plan be submitted to and accepted by the SMO prior to the placement of mass concrete on a given project. Therefo re, the FDOT may decide not to accept a given concrete mixture design as provided by the contractor. A survey of mixes within the FDOT SMO database was performed to determine the typical cementitious material type and content mass concrete mix design. Th e database contains mass concrete mix design information for mixes used in the state of Florida between 1980 and the present. Table 2 -4 provides a brief summary of 32 typical mix designs used for mass concrete The mix designs include the use of Type II cement, replacement of Portland cement with Grade 100, granulated Blast Furnace Slag, or Type F fly ash. A search of the FDOT database revealed that

PAGE 38

38 the FDOT has not accepted a mass concrete mix design which does not incorporate supplementary cementitious materials since 1980. Type II Portland cement produces less heat during the process of hydration than Type I cement (ASTM C 150) Supplementary cementitious materials (as a replacement for Portland cement), generate less heat than concrete mixes which utilize Portland cement alone (Malhotra and Mehta, 1996). The physical and chemical attributes of supplementary cementitious materials are discussed in further detail in subsequent chapters. Table 2 -4 Typical c ementitous c ontent u sed for mass c oncrete Name FDOT Class Cement Type Cement (lbs) Fly Ash (lbs) Blast Furnace Slag (lbs) Fly Ash % Blast Furnace Slag % Mix #1 IV II 610 134 0 18.01 0 Mix # 2 IV II 600 132 0 18.03 0 Mix # 3 IV II 359 83 0 18.78 0 Mix # 4 IV I 625 150 0 19.35 0 Mix # 5 IV II 3 14 77 0 19.69 0 Mix # 6 IV II 584 146 0 20 0 Mix # 7 IV II 576 144 0 20 0 Mix #8 II II 457 115 0 20.1 0 Mix #9 II II 344 83 0 19.44 0 Mix #10 II II 329 0 115 0 25.9 Mix #11 IV II 500 200 0 28.57 0 Mix #12 IV Drilled Shaft II 523 257 0 32.95 0 Mix #13 IV II 500 260 0 34.21 0 Mix #14 IV II 428 230 0 34.95 0 Mix #15 IV Drilled Shaft II 478 257 0 34.97 0 Mix # 16 II II 500 270 0 35.06 0 Mix # 17 IV II 438 243 0 35.68 0 Mix # 18 IV II 438 292 0 40 0 Mix # 19 IV II 450 383 0 45.98 0 Mix #20 IV Drilled Sh aft II 440 390 0 46.99 0 Mix #21 IV II 346 320 0 48.05 0 Mix #22 II II 296 274 0 48.07 0 Mix #23 II II 296 274 0 48.07 0

PAGE 39

39 Table 2 4. Continued. Name FDOT Class Cement Type Cement (lbs) Fly Ash (lbs) Blast Furnace Slag (lbs) Fly Ash % Blast Furnace Slag % Mix #24 IV II 336 329 0 49.47 0 Mix #25 II II 196 0 196 0 50 Mix #26 IV II 450 450 0 50 0 Mix #27 IV II 330 0 330 0 50 Mix #30 V II 370 0 520 0 58.43 Mix #31 IV Drilled Shaft I/II 290 0 436 0 60.06 Mix #32 IV II 200 0 460 0 69.7 Tensil e Stresses in Mass Concrete As stated earlier, tensile stresses are the primary cause of failure of mass concrete structures. Tensile stresses in mass concrete typically evolve from two primary sources: thermal gradients throughout the concrete structure a nd delayed ettringite formation. Tensile S tresses D ue to T hermal G radients Thermal gradients are primarily induced by heat loss from the outer surface of the mass concrete structure. As cement hydrates, it produces heat, increasing the temperature of the c oncrete. Should the outer surface of the concrete be exposed to external temperatures lower than those developed inside the structure by hydration, a temperature gradient will be created. Large differences in temperature will lead to thermal stresses that can result in cracking. This phenomenon is intensified by the use of cements high in C3S and/or C3A, as these compounds are responsible for the majority of early heat development. It can also be intensified by cements with higher fineness, resulting in fas ter reaction due to increased surface area (Price, 1974; Poole, 2004; Neville, 1995; Woods et. al., 1933; Larsen, 1991; Higginson, 1970) As shown in table 2-3 those DOTs which have requirements for mass concrete also specify limitations on the maximum al lowable temperature gradients for early age mass concrete. Most of these DOTs require a maximum temperature differential of 35 F, which has

PAGE 40

40 become the most commonly used temperature differential for mass concrete applications today. This approach was base d on the data collected from a project involving the construction of small concrete dams in England more than 50 years ago (Fitzgibbon, 1976; Gajda, 2007). Although no readily available laboratory data confirms the suitability of a maximum temperature diff erential of 35 F, it is nonetheless still required in many mass concrete specification documents today. The exact magnitude of the temperature gradient depends on a number of factors including the initial concrete temperature, the ambient temperature ar ound the structure, and the thermal properties of the concrete itself. A number of early age concrete properties affect t he thermal attributes of concrete i ncluding heat development, tensile strength development, coefficient of thermal expansion, thermal diffusivity, resultant strain gradients, temperature gradients, and the chemistry of the cement paste. Supplementary cementitious materials can dramatically reduce the amount of heat generated. One of the principal difficulties in predicting cracking of co ncrete during hydration is that the tensile stresses develop before the concrete has reached its ultimate strength. U nder certain conditions such stresses can develop while portions of the concrete placement are still in the plastic state, resulting in p lastic shrinkage cracking. To model mass concrete performance throughout the early hydration stage (wherein heat generation is maximized), it is first necessary to characterize the development of stress -strain behavior of the concrete. Stresses due to the rmal gradients within a mass structure change over time, as does the stress -strain behavior of the concrete. T he relative magnitude of these two variables is important when considering the potential for cracking. In at least one instance, published researc h has claimed that thermal gradients can be mitigated by using faster hydrating cements, asserting that tensile strength develops faster than the thermally induced tensile stresses (Lindstrom, 2003).

PAGE 41

41 Though intriguing as a possibility, this research did not consider the rate of development of tensile elastic modulus. While strength could very well develop more rapidly than the thermal stresses, stiffness may possibly increase at an even faster pace. This would result in a lack of strain capacity that would lead to cracking due to thermal expansion. T o account for the effect of these interrelated factors, the stress -strain behavior of concretes at different ages and temperatures must be investigated. Additionally, past research has focused on the compressive strength of the concrete when determining its resistance to applied stresses. Though compressive strength is the primary design parameter in mass concrete, thermal cracking is inherently a tensile phenomenon. Thus the tensile strength properties are criti cally important in the modeling of mass concrete behavior. Unfortunately, tensile strength of concrete can be very difficult to measure with traditional techniques that are either extremely difficult to perform (direct tension) or insensitive to early micr o-cracking (splitting tension) (Bremner et. al, 1998; Boyd and Mindess, 2002; Boyd and Mindess, 2004) An alternative method for determining tensile properties is to test the tensile modulus of elasticity using flexural beam testing. Tensile Stresses Due t o D elayed Ettringite F ormation Delayed ettringite formation (DEF) is a form of internal sulfate attack on concrete that can lead to severe cracking and damage and eventual material failure. During the hydration of Portland cement under normal conditions, e ttringite forms as a result of the chemical reaction between C3A and gypsum. The ettringite then further reacts with the remaining C3A to produce monosulfoaluminate, a relatively inert compound. Gypsum is added to cement in order to force this ettringite f ormation and prevent flash set of the C3A. As long as the ettringite forms while the concrete is still plastic, it is essentially harmless.

PAGE 42

42 Under high temperatures, however, the ettringite is destroyed and its component s sulfate and aluminate are absorbed by the calcium -silicate hydrate (C -SH) in the hydrated cement paste. After cooling, the sulfate is released and again becomes available to form ettringite. Ettringite is an especially crystalline compound with a high aspect ratio that can form long needle like crystals. Once nucleation has occurred, ettringite crystals tend to grow lengthwise. Even when the tips of the crystals encounter solid material the growth continues resulting in very high localized stresses that lead to expansion and cracking of t he hydrated cement paste matrix. The exact temperature required for the formation of DEF has been extensively debated. The majority of available research reports 158 F (70 C) as the minimum temperature necessary for the formation of DEF (Lawrence, 1998; Taylor, 1997; Gajda, 2007; Drim a las, 2004). Most of the DOT specifications listed in table 2.2 require a maximum curing temperature of 160F for mass concrete structures in an effort to avoid the onset of DEF. However, t here are specific conditions necess ary for the initiation of DEF. Even in the presence of high temperatures it only occurs within cements with large amounts of SO3 (4%), alkalis, or MgO, and in the presence of moisture. However, the exact effect of cement chemistries is not fully understood (Taylor, 1997). The combination of high curing temperatures and moist environments is most common in two situations: pre -cast concrete exposed to steam curing, and massive concrete members in a damp environment (drilled shaft piles, foundations, abutment s, etc.). To avoid DEF in mass concrete, it is necessary to prevent the temperature rise in the concrete from exceeding the threshold value at which ettringite nucleation occurs. The exact threshold temperature is affected by several factors though little information is currently available concerning these parameters for Florida concretes.

PAGE 43

43 DEF can be difficult to diagnose especially at the early ages necessary for this research. The minute scale of the crystals makes them impossible to find using optical microscopy and their formation does not change the chemical composition of the bulk hydrated cement paste, thereby making traditional chemical analysis inadequate. The ettringite crystals themselves do exhibit a unique chemical makeup from the rest of the paste, but they form due to chemical realignment in the paste and thus any reasonable sample size will exhibit the same chemical makeup as it did before DEF formation. Identification of ettringite alone is not sufficient to indicate DEF, since ettringite does form during normal hydration and can be left over if the C3A/gypsum balance of the cement is not perfect. Small amounts of late -forming DEF are also harmless and could actually be beneficial since they will initially fill in any voids in the hydrated cement paste, thus densifying the microstructure and improving strength and permeability. This is commonly called Secondary Ettringite formation (Skalny et. al, 2002). Only when sufficient ettringite forms to expand beyond the allowable void space does it become a problem. Such incidents are evidenced by cracks in the microstructure initiating from the ettringite crystals. A rigorous literature search was performed in an effort to properly address the relevant issues concerning DEF and mass concrete struc tures in Florida. As previously stated, DEF is a form of internal sulfate attack which is a result of the high heat of hydration coupled with sulfates in the cement paste and a relatively high C3A content of the cement (Ghorab, 2002). The FDOT permits the use of supplementary cementitious materials to replace Portland cement in the construction of massive elements in which the core temperature of the concrete is expected to exceed 165F FDOT specification allows use of 18%50% Portland cement replacement o f fly ash and 50%70% replacement of ground blast furnace slag to ensure lower

PAGE 44

44 heats of hydration (FDOT, 2007). Currently, all of mass concrete construction in the state of Florida utilizes Class F fly ash or slag as a major constituent of the cementitious material in the concrete mix design. A significant portion of the literature review was dedicated to the effect of fly ash and slag on DEF and sulfate attack of concrete materials and structures. It has been shown that using 25% fly ash or 25% slag replac ement in concrete mitigates the physical mechanisms of DEF (Ramlochan et.al, 2003). Similar findings have been reported by other researchers (Lane and Ozyildrum, 1999; Miller and Conway, 2003; Thomas et. al, 2008). Despite the fact that the separate physi ochemical mechanisms that cause internal sulfate attack (DEF) and external sulfate attack are different, the expansive nature of the resultant product (ettringite) is the same for both. Consequently, the literature review included the study of external sul fate attack on ordinary and mass concrete and the effect of fly ash and slag on the mitigation of sulfate attack. The use of fly ash and blast furnace slag in making sulfate -resisting concrete has frequently been reported (Bhatty and Taylor, 2006). Othe r researchers have reported similar findings (Skalny et. al, 2002; Malhotra and Mehta, 1996; Schlorholtz and Bergeson, 1993). Research has also been performed to evaluate massive concrete structures which have shown deterioration due to sulfate attack. S everal bridge structures in Europe were investigated to determine the nature of sulfate attack and the possible mechanisms of the deterioration. The study determined that the most likely cause of initial cracking was thermal expansion. The sulfate attack which took place after the initial cracking was more likely to be due to an external sulfate source rather than from within (Divet, L., and Pavione, 2002). T he initial cracking of the structure w as attributed to a severe temperature differential between the outer concrete strata and the internal core. Upon the initiation of the initial cracking, it was determined that the presence

PAGE 45

45 of moisture and sulfate from the environment, rather than internal sulfate attack or DEF, contributed to years of deterioration (Divet, L., and Pavione, 2002). More recently, DEF has been found in a mass concrete bridge structure that is currently in service. The bridge structure was diagnosed with alkali silica reaction (ASR) in some portions, DEF in others, and some of the struc ture was considered to be free of significant damage. This is the first well -documented case of DEF occurring in a concrete structure in which the temperature rise of the concrete was due solely to internal hydration (i.e. the structure was not steam cure d). Interestingly, the structure that sustained damage due to ASR and DEF was concrete which did not contain supplementary cementitious materials. The portion of the structure that was free of significant damage contained fl ya sh. (Thomas et.al, 2008) Re search has been performed investigating the use of granulated blast furnace slag as a partial replacement of Portland cement in mortars exposed to high temperatures. The results of this research indicate that mortars which use 30% and 50% granulated blast furnace slag experience damage due to expansion or DEF when exposed to curing temperatures of 208F (98C) (Kelham, 2002). Thus, it is likely that mass concrete containing large replacements of supplementary cementitious materials approved per the FDOT st andard specification for Road and Bridge Construction will not be subject to DEF due to the following reasons: The temperatures necessary to cause the dissolution DEF at early ages are less likely within concrete structures containing supplementary materia ls The addition of supplementary materials mitigates the damage produced by DEF DEF in massive structures containing supplementary cementitious materials is less likely as the near surface temperature is less than the core temperature. However, the core t emperature is less likely to be exposed to moisture due to the relatively large amount of concrete cover. As a result of the literature review as well as interviews conducted with FDOT personnel and several onsite visits, it was determined that the factor s most relevant to cracking in massive

PAGE 46

46 structures are thermal stresses induced by a thermal gradients. Therefore, the DEF study was altered to include a rigorous study of the components which are most likely to contribute to large thermal gradients in mass ive concrete structures. The major components are as follows: High internal heats of hydration caused by the concrete mix design. Poor insulation of mass concrete at early ages. M ix designs which utilize too much cementitious material. The use of cements w ith high heats of hydration. Along with FDOT personnel interviews, a review of the current practice of the development of thermal control plans for mass concrete construction was conducted to determine which issues contribute to thermal gradients in mass concrete. The majority of FDOT construction using mass concrete involves bridge components in severely aggressive environments. Per the FDOT standard specification, Type II Portland cement must be used on projects constructed in aggressive environments. Consequently, this research will investigate the chemical and physical properties of concrete mixes using primarily Type II cement enhanced with Type F fly ash and/or Ground Blast Furnace Slag. Concrete with Type I Portland cement will also be studied for co mparison purposes.

PAGE 47

47 CHAPTER 3 ISOTHERMAL CONDUCTION CALORIMETRY TESTING Introduction Isothermal conduction calorimetry has gained much popularity throughout the building and construction industry over the past two decades. As a result, ASTM C 1679, the s tandard practice for measuring hydration kinetics of hydraulic cementitious mixture s using isothermal calorimetry has been developed for measuring the relative differences in hydration kinetics of hydraulic cementitious mixture s (ASTM, 2008b). Additionally ASTM is currently in the process of writing a standardized test method for the measurement of heat of hydration of cement with heat conduction calorimetry (Robbins, 2007). Isothermal conduction calorimetry measures the heat evolved from a relatively sm all specimen over time. Therefore, this testing method is very useful for the testing of the cementitious components to determine the heat evolved or the heat of hydration of the materials. Currently, there is no analysis method which predicts the heat ris e in massive concrete structures that incorporates the results provided by isothermal conduction calorimetry for input parameters. There is a model proposed by Poole et. al (2007) and Riding et. al ( 2007) that utilize s semi adiab atic calorimetry for the p rediction of temperature rise of mass concrete S ome researchers state that the semi adiabatic calorimetry is not accurate for the measurement of the heat of hydration / heat production at early ages (Wadso, 2003, Robbins, 2007). ASTM has not standardi ze d either test method to date. H owever, a standard test method for the use of isothermal calorimetry to measure heat evolution of cementitious materials will more likely be available before a standardized test method for measuring heat evolution / temperatu re rise of concrete using semi adiabatic testing of concrete becomes available. The reason for the potential publication of a standardized test method for the isothermal calorimetry

PAGE 48

48 test is that there is currently a draft test method before ASTM committee C.1.09. (Robbins, 2007) However, the semi adiabatic test method has not been presented to a standardization committee for publication. Isothermal calorimetry provides a direct measurement of heat generated for cementitious materials. Several r esearchers h ave reported the use of cement chemistry to determine apparent activation energy and total heat available in a cementitious constituent as provided by XRF or XRD analysis (Ma et. al, 1994, Poole et. al, 2007, Schindler 2004, Schindler and Folliard, 2005) Chapter 4 provide s a detailed analysis of the determination of apparent activation energy using several methods, including isothermal calorimetry data. The original equations used for the conversion of chemical oxides to the cement phases were developed b y Bogue in the 1940s (Bogue, 1947). R ecently it has been reported that the chemical data obtained via XRF have relative bias and potential errors (Stutzman, 2007). Thus, the use of chemical analysis for the prediction of heat of hydration could have erro rs accordingly. Since isothermal calormetry provides a direct measurement of heat evolution from cementitious materials, the potential errors per chemical analysis can be avoided accordingly. Summary of Test Method An isothermal heat conduction calorimet er consists of a constant temperature heat sink which two heat -flow sensors and sample holders are attached. One heat -flux sensor and sample holder contains the test sample ( with cementitious component). The other heat -flow sensor is composed of a referenc e cell containing a sample that evolves no heat. The heat of hydration released by the hydrating cementitious sample is passed across the sensor to the heat sink. The output measured by the calorimeter is the difference in heat flow (thermal power) between the sample cell and the reference cell. The physical measurement of heat flux is due to a small temperature gradient that develops from the sample side to the heat -sink side. The temperature

PAGE 49

49 d ifference is so small that for practical purposes, the samp le is at a constant temperature isothermal). The output from the heat -flow sensor is an mV signal that is proportional to the thermal power from the sample. This output must be calibrated to a known thermal power. In this method, this is accomplished by measur ements on a sample that emits a constant and known power level. The integral of the thermal power over the time of the test is the heat of hydration. Equipment and Procedure The FDOT SMO houses an isothermal conduction calorimeter made by Thermometric AB c alled the TAM -Air isothermal calorimeter as shown in Figure 3 1. The calorimeter has eight channels which allows for the monitoring of eight individual specimens at a time and an operating temperature range of the calorimeter is from 5 C to 60 C. A detai led operating procedure for the isothermal calorimeter is provided in Appendix A. Figure 3 1. Isothermal conduction calorimeter test set up M ix ing Procedure There are two acceptable methods for the introduction of the water into dry cementitious materi als for isothermal conduction calorimetry testing. The methods are referred to as external and internal mix ing.

PAGE 50

50 External M ix ing External mix ing is the process by which the water and cem entitious materials are mix ed in a device, typically a high-speed b lender outside the calorimeter. Researchers often use this method because it requires less time for specimen preparation and does not require an internal mix ing calorimeter cell which is relatively costly. The main limitation of the use of the external mi x ing procedure for the measurement of heat of hydration of cemenitious materials is that this procedure requires that the total heat of hydration calculation be corrected for the heat evolved prior to the introduction of the cementious specimen into the calorimeter (Poole, 2007). Due to recent developments of the (unpublished) isothermal calorimetry test method the precision and bias statement regarding the corrections needed for the external mix ing procedure have not been developed. Internal M ix ing The internal mix ing procedure involves the use of a calorimeter cell filled with water attached to an ampoule filled with dry cemen t as show in Figure 32. The internal mix ing procedure is provided in detail in A ppendix A. Figure 32. M ix ing cell and ampo ule used for internal mix ing procedure

PAGE 51

51 There are several main advantages for using the internal mix ing procedure over the external mix ing procedure. The intern a l mix ing procedure requires the calorimeter cell be placed inside the calorimeter to establish t he internal temperature conditions prior to initiation of mix ing Unlike the external procedure, the specimen is known to be at the temperature of the system prior to mix ing. Another difference between the internal mixing and external mixing procedure is t hat the internal mixing procedure allows for the more precise measurement of the mass of the total cementitious product for mixing. The external mixing procedure tends to require a relatively swift transfer of material (less than 5 minutes) from the mixer to the ampoule at the time of mixing. The rapid nature of the transfer d oes not facilitate the accurate proportioning of the mix ed cementitious material. One other potential variable in the mix ing procedure can be attributed to the nature of the externa l m ix er itself. As proposed by Poole (2007) a high -speed blender should be used to mix the materials, which could potentially introduce heat due to the friction of the mix ing procedure. Another variable can be attributed to external mix ing procedure is the potential heat transfer from the operator to the specimen itself. Since the external mix ing procedure has more potential variables for the alteration of the heat of the sample without quantification, it was decided to use the internal mix ing procedure for the isothermal cement calorimetry in this research. Most research to date using the Tam air calorimeter (Poole, 2005, Wadso, 2003,) reports the use of isothermal calo rimetry data with the use of external mix ing procedure s However, internal mix ing has been reported for research on cementitious materials as well (Evju, 2003). Isothermal Calorimetry Testing Methodology Isothermal conduction calorimetry has two main applications for the testing of cementitious materials determination of heat rise, and deter mination of activation energy. In this chapter, the isothermal calorimetry testing result s of a single Portland cement tested at

PAGE 52

52 various temperatures and water to cementitious ratios are reported to discuss the reliability of internal mix ing procedure with regard to this test method, for the determination of heat rise. In chapter 4, the results of isothermal calorimetry testing for the determination of activation energy of cementitious mixture s is reported. To determine the validity of the internal mix ing procedure for the measurement of total energy rise, t esting was performed on two Portland cement mix es at temperatur es of 15, 23, 38, 49 and 60 C, and water to cement ratios of 0.50 and 0.40. The Tam air isothermal conduction calorimeter has an operating temperature range from 5C to 60C. Isothermal Calorimetry Testing Results The results from the isothermal calorimetry testing are typically reported in terms of power per unit mass (W/g) vs. time and energy per unit mass (J/g) vs. time. Figure 33 show s the resultant calorimetric curves for the 0.50 w/c ratio mix with regard to the power for each isothermal temperature for the first 180 hours The curves indicate that the larger temperatures achieve larger total power at earlier ages than the lower temp eratures. This result is consistent with the findings of others (Ma et. al, 1994, Poole et. al, 2007, Broda et.al, 2002, Wadso, 2003). The results are also consistent with the typical characteristics of the hydration of Portland cement based materials (Meh ta, 1986). As discussed in A ppendix A, the total heat of hydration of the sample is calculated by integrating the power/g versus time data over the time interval of the test as shown in equation 3 1 (Poole, 2007) Thus, the energy is calculated directly f rom the power. (31)

PAGE 53

53 Figure 3 4 shows the resultant calorimetric curves for the 0.50 w/c ratio mix with regard to the energy for each isothermal temperature for the first 180 hours. As to be expected, the curves indicate that the higher temperatures evolve energy at faster rates than those of the lower temperatures. However, the Portland cement tested at 60C, did not obtain a higher energy as compared with the specimens at lower temperatures. Figure 3 3. Power vs. time for 0.50 w/c Portland cement at various temperatures Figure 34. Energy vs. time for 0.50 w/c Portland cement at various temperatures

PAGE 54

54 This test result can be attributed to a limitation of the internal mix ing procedure. Upon the completion of the test, at the moment the specimen was removed from the calorimeter, the ampoule was observed to have condensed water droplets distributed on the inside. This phenomenon is most likely a result of incomplete hydration due to the evaporation of the mix ing water from the matrix which deposited onto the glass. The resultant hydration product appeared to be an incompletely hydrated cementitious product. Figure 35 illustrates a side -by -side comparison of a fully hydrated specimen tested at 23C (left) with a partially hydrated specimen tested at 6 0C (right). Figure 35. Fully hydrated specimen (23C) vs. partially hydrated specimen (60C) As a result of the heat of the incomplete hydration of the cementitious product, as indicated by the presence of the water droplets which collected on the am poule, the resultant power curve actually obtained negative readings at later ages (approximately 140 hours). This results in the corresponding energy curve acquiring a slightly lower energy value. This phenomenon is consistent with the conditions observed upon the removal of the ampoule from the calorimeter. The water required energy input to maintain the isothermal temperature of 60 C which resulted in a measured energy loss. The same phenomenon can be observed for the cementitious specimens mix ed with a lower w/c at 49C as shown in Figure 36.

PAGE 55

55 Figure 36. Energy vs. time for 0.40 w/c Portland cement at various temperatures The specimen tested at 49C for the 0.40 w/c does not behave exactly like that at 60C as the results indicate that the 49C spec imen does not lose energy at later ages. However, the data presented in Figure 36 shows the first 72 hours hydration. The testing regimen was shortened for the investigation of the lower w/c mixture as it was performed solely in an effort to determine an y potential effects of w/c on the internal mix ing procedure and calorimetric results. Summary of Results The testing of the various Portland cement pastes with isothermal calorimetry resulted in the following conclusions. Isothermal calorimetry preforme d on mix es at higher temperatures provides larger rates of energy rise with respect to time. This result is consistent with available literature. (Wadso, 2003, Ma et.al, 1994, Schindler 2003, 2004, Broda et.al, 2002 and Poole, 2007) The use of the interna l mix ing procedure for the isothermal calorimetry testing is a viable method for the measurement of early age energy rise. The isothermal calorimetry using the internal mix ing procedure resulted in partially hydrated specimens when tested at temperatures above 49C. Therefore, i sothermal calorimetry testing for the specimens in this project would be limited to temperatures of 49C and below.

PAGE 56

56 Due to some of the limitations of the internal mix ing procedure, t he use of a 0.50 w/c mixture for the cementitiou s would be established for all of the concre te mix designs in this project. This was decided because the mix es tested at a w/c below 0.50 did not provide a typical energy profile at higher temperatures.

PAGE 57

57 CHAPTER 4 MATURITY AND EQUIVAL ENT AGE As previously stated, the hydration of cement in concrete (an exothermic process which results in the evolution of heat) results in the rise of concrete temperature. The resultant product of the hydration of Portland cement paste is the hardened cement paste matrix which, despite the fact that the paste is initially fluid, becomes hard and strong as time passes (Young et. al, 1998). This chapter discusses the maturity concept used to quantify the relative or equivalent age of concrete with respect to time -temperature history. The investigation of various maturity and equivalent age functions are investigated in this chapter. The maturity/equivalent age functions established in this chapter are then used in each of the subsequent chapters for the analysis of the respec tive physical or thermal properties. Maturity Concept The maturity method is one technique used for the estimation of the development of strength of hardened concrete. Like many other chemical reactions, the rate of the reaction (hydration) of Portland c ement is influenced by the temperature of the reactants. Thus, the increase in concrete temperature will result in a corresponding increase in the rate of the hydration reaction. The first maturity functions for concrete were developed in the late 1940s by McIntosh, Nurse and Saul (Carino, 2004). The term maturity and the Nurse Saul maturity function originated from this work and the first widely accepted function was accepted by industry for the use of quantification of time temperature history of concrete. However, it was Saul that stated that under certain conditions, a given mixture of concrete with the same maturity will have the same strength (Tank, 1988). As shown in Figure 4 1, the maturity of concrete can be estimated by simply determining the area under the time temperature curve using the following equation:

PAGE 58

58 Figure 4 1. Nurse -Saul maturity function (Carino, 2004) tTTM )(0 (41) Where: M = maturity at age t T = average temperature of the concrete during the time i nterval t To = Datum temperature Per equation 4 1, the maturity of the concrete is quantified in units of time (hours) and temperature (C).e.g., (C -hr). The datum temperature is considered to be the lowest temperature at which the hydration reaction will occur (typically below 0C ) The standard practice of estimating concrete strength by the maturity method (ASTM C1074) describes the standardized procedure for calculating maturity and the datum temperature for concrete (ASTM, 2004c). Equivalent Age Concept The concept of equivalent age expresses the time temperature history of concrete in units of time rather than using temperature time units as described by the maturity method. The equivalent age of a concrete is defined as the number of days or h ours at a specified standard temperature required to produce a maturity value equal to the value achieved by a curing period at temperatures different from the specified temperature. According to ASTM C1074, the actual

PAGE 59

59 age of the concrete is transformed to its equivalent age at specified temperatures by means of a maturity function (Tank, 1988 p.33). There are a number of different equations that can be used to calculate the equivalent age of concrete. For instance, it is possible to use the maturity function per equation 4 1 to create an equivalent age as follows (Tank, 1988): )( )(0 0 0TT tTT tS t e (42) Where: te = equivalent age at a standard temperature Ts(C) t = time (hours) Arrhenius Equation and Activation Energy Other equations have been developed for determining equivalent age (based on temperature and time) which have been used for predicting the rate of strength gain in concrete. The Arrhenius equation is nonlinear and describes temperature dependent reactions as follows (Glasstone et. al, 1941): RT EaAe k (43) Where: k = the rate of heat evolution (W) A= proportionality constant (W) Ea = Activation Energy (J/mol) R = Universal Gas Constant = 8.314 (J/mol/K) T= Temperature (K) Activation energy is defined as the heat of activation or energy of activation of the reaction; it represents the energy that a molecule in the initial state of the process must acquire

PAGE 60

60 before it can take place in the reaction whether it be chemical or physical (Glasstone et.al 1941 p2) Unlike the Nurse Saul method, the Arrhenius equation requires the determination of the apparent activation energy (Ea) for the calculation of maturity and equivalent age functions. Ea represents the energy that the molecule in the initial state of t he reaction process must acquire before it can participate in the (hydration) reaction. The experimental Ea can be obtained from the linear plot of reaction rate vs. the inverse of the temperature (Glasstone et. al, 1941). More specifically, the experiment al activation energy is determined by multiplying the negative value of the universal gas constant (R) by the slope of the best fit line through ln(k) versus 1/(T*R) Glasstone discusses that the experimental activation energy is obtained via the best f it line procedure, which is slightly different from the activation energy or the heat of activation, depending upon the reactive species, energy liberated from reaction, and rate of reaction (Glasstone et. al, 1941). The term activation energy for the purposes of this research is interchangeable with experimental activation energy as defined by Glasstone. The obtained value of activation energy is used to characterize the reaction rate of cementitious materials at various temperatures (Poole, 2007 p7) The Arrhenius equation in the form presented in equation 4 3 is not suitable for the calculation of equivalent age. The concept of using the Arrhenius equation for the description of the hydration of Portland cement and concrete materials was first sugg ested by Freiesleben Hansen, and Pederson in the late 1970s (Freiesleben, et.al, 1977). The equation proposed by Freiesleben Hansen, and Pederson, is commonly referred to as the Arrhenieus equation within the Portland cement concrete industry, and is sp ecifically used for the application of the equivalent age concept with regard to concrete maturity:

PAGE 61

61 t eTtt TTR E rec r 0 273 1 273 1 (44) Where: te(Tr)= equivalent age at the reference curing temperature Tc = average temperature of the concrete during t he time interval t Tr = reference temperature (typically 22.8C / 295.8K) E = activation energy, J/mol R = universal gas constant, 8.3144 J/(mol K) ASTM C 1074 utilizes the Arrhenius equation as presented in equation 4 4 to determine equivalent age for concrete materials as an alternative to the Nurse -Saul maturity function (ASTM 1074 6.3) with some slight differences in the notation of the equation. t ett TT Q es a 0 273 1 273 1 (45) Where: te = equivalent age at a specified temperature Ts, days or hours Q = activation energy divided by the universal gas constant Ta = average temperature (C) of the concrete during the time interval t Tr = Specified Temperature, (typically 22.8C / 295.8K) The equivalent age concept has been researched and applied to Portland cement concrete materials by many researchers primarily for the estimation of the concrete strength properties (Carino, 2 002, Carino and Tank 1992, Carino and Lew, 1982, Carino et. al, 1982, Volz et. al, 1980, Voigt et, al, 2006, Tank, 1988). Activation Energy Determination via Compressive Strength On the day of concrete delivery for each large scale block specimen, Mr. Joseph Fitzgerald, concrete mix design specialist from the FDOT State Materials office, visited the Florida Rock Industries r eady -mix plant in Gainesville, Florida where the concrete was batched.

PAGE 62

62 Mr. Fitzgerald obtained the dry Portland cement, sand, coarse aggregate and relevant supplementary cementitious materials that were used in the concrete delivered to the State Materials Office for laboratory specimen creation. This material was obtained for the creation of laboratory specimens at a later date. Tabl e 4 1. Mortar c ube m ixture d esigns mix Name Cement (g) Slag (g) Fly Ash (g) w/c Water (g) Sand (g) Lab mix (100% Portland Cement) 4086 0 0 0.5 2043 6570 M ix 1 (100% Portland Cement) 4086 0 0 0.5 2043 6084 M ix 2 (50% Portland 50% Sla g) 2046 2046 0 0.5 2046 6084 M ix 3 (65% Portland 35% Slag) 2658 0 1428 0.5 2046 6066 M ix 4 (50 30 20 Blend) 2046 1224 816 0.5 2046 6552 Some of this material was used to create mortar cubes with the same mixture proportions as the lar ge -scale block specimens in accordance with ASTM C 1074 for determination of Ea. As shown in Table 4 1, the mortar cube mixes were proportioned with the appropriate Portland cement, SCM (supplementary cement materials) (where applicable), 0.50 w/cm, and t he appropriate amount of sand per ASTM C109 for standard cube specimens. The specimen group designated lab mix was created to perform laboratory tests for isothermal calorimetry testing. Unlike the other specimen groups illustrated in Table 4 1, the lab mix does not correspond to the mix used in the large block specimen but has the same cement paste composition.

PAGE 63

63 Apparent Activation Energy Determination as Per ASTM C 1074 ASTM C 1074 requires at least three sets of mortar cubes be created and exposed to three different isothermal curing temperatures for the calculation of Ea. Temperatures of 8C, 23C, and 38C were chosen as the isothermal temperature groups for this research. The temperature of 23C is considered to be standard curing temperature T he other specimen groups were stored at 15C above and below the standard curing temperature. The mortar cubes were prepared in accordance with ASTM C 109 However, to ensure isothermal conditions for the entire history of the specimens, the cement, mortar and water and specimen molds were stored in an environmental chamber at the isothermal temperature of 8C or 38C for 12 hours prior to the fabrication of the specimens. Upon the completion of the specimen fabrication, the molds were immediately returned to their respective environmental chambers for 24 hours, after which the specimens were de molded and stored in a water bath at the appropriate temperature until compressive strength testing was performed. The calculation of Ea per ASTM C 1074 requires the determination of several parameters per equation 4 6 as follows: )(1 ) (0 0ttk ttk SSu (46) Where: S = average cube strength at age t (psi) t = test age (hours) Su = limiting strength or u ltimate strength (psi) t0 = age when strength development is assumed to begin (hours) k = rate constant. Per ASTM C 1074 A1.1.8.1, if the user has access to a computer program that will permit the fitting of a general equation to a set of data The computer program will calculate the best

PAGE 64

64 fit values for Su, t0 and k. (ASTM, 2004c). This can be performed by creating a simple spreadsheet in a program such as Microsoft Excel into which the user programs equation 4 4 and employs regression and the Excel Solver to find the best fit solution. Appendix B outlines the methodology for the curve fitting procedure. Figure 4 2 is a graphical representation of compressive strength versus time for both actual values and the predicted values per equation 4 6 (hyperbolic / ASTM 1074) for mix #1. The results of compress ive strength versus time for each mortar mixture can be found in Appendix C. Table 4 2 is a summary of the strength development parameters computed for each mortar mixture. Figure 4 2. Compressive strength of mortar cubes vs. time for each temperature Th e observed data regarding compressive strength of cubes versus time compare relatively well with the calculated data as per ASTM C1074A1.1.8.1. Table 4 2 is a data summary of the curve fit / strength development parameters for each of the mortar mixes. Th e apparent activation energy is calculated by determining the best fit straight line from the natural logarithm of the k values versus the inverse of the absolute temperature T he negative slope of the line is

PAGE 65

65 multiplied by the universal gas constant R (8.3144 J/mol -K) to determine Ea, as shown in Figure 4 4. T able 4 -2. Strength d evelopment p arameters for each c oncrete m ixture per ASTM 1074 Mix Name Temp (C) Su (psi) k T o (hr) R 2 Ea (J/mol) Mix 1 (100% Portland Cement) 8 9330 0.004 17.6 0.9950 35642 23 9828 0.008 5.5 0.9998 38 8327 0.019 1.9 0.9944 Mix 2 (50% Portland 50% Slag) 8 7064 0.002 15.8 0.9970 33688 23 8979 0.005 7.9 0.9997 38 11446 0.008 4.0 0.9981 Mix 3 (65% Portland 35% Slag) 8 6300 0.002 15.0 0.9747 25757 23 9018 0.0 04 7.5 0.9991 38 9290 0.005 3.2 0.9578 Mix 4 (5030 20 Blend) 8 5529 0.002 9.0 0.9976 30013 23 7152 0.005 7.7 0.9990 38 11437 0.006 3.3 0.9939 Figure 4 -3 Activation energy calculation for hyperbolic method Activation Energy Determination Via Exponential Method The determination of Ea via the exponential method is performed using the same procedure for the ASTM C 1074 method. The difference between the method standardized by ASTM and the exponential method is th at the exponential method ut ilizes a n exponential instead

PAGE 66

66 of a hyperbolic equation to fit the strength versus time data. The calculation of activation energy per the exponential method requires the determination of several curve fit parameters within each equation per equation 4 7 as follows (Schindler, 2002): t SSs uexp (47) Where: S = average cube strength at age t (psi) t = test age (hours) Su = limiting strength or u ltimate strength (psi) = shape constant = time constant for strength prediction (hours) Equation 4 7 is used to create a best fit equation for the evolution of strength vs. time in a manner similar to the method outlined in ASTM C1074 (using a solver equation), which is then used to calculate the Ea of the cementitious material. The primary difference between the methods is the general form of the equationhyperbolic vs. exponential. Schindler, (2002) states : If the experimental data are available at temperatures other than the reference temperature, the best fit hydration parameters ( ) at each of the temperatures can be determined. When the exponential equation is used, it is assumed that the slope parameter is independent from the curing temperature (Schindler, 2002 p.205). To create the exponential curve fit relationship, Schindle r used data provided by Kjellesten and Detwiler (1993) to Ea based on degree of hydration rather than strength development incorporated into equation (4 8). ts uexp (48) Where: = degree of hydration t = test age (hours) u = ultimate degree of hydration = shape constant = time constant for strength prediction (hours)

PAGE 67

67 The following procedure was used for the determination for the Ea using hydration parameters per Schindlers research : The best fit degree of hydration curve for the experimental data was obtained for the reference temperature (23C). The following parameters were determined : = hydration time parameter at the reference temperature (hours) = hydration shape parameter u = ultimate degree of hydration The hydrat ion shape parameter ( ) and the ultimate degree of hydration ( u) were determined at the reference temperature and were used as constants in the degree of hydration curves at the other test temperatures. The best -fit hydration time parameter s ( ) at all t est temperatures are determined using a solver / curve fitting software. The Arrhenius plot is constructed, by plotting the natural logarithm of the hydration parameters ( ) for each temperature versus the inverse of the absolute curing temperature. (Sc hindler, 2002 p.207) Figure 4 7 is a graphical representation of the compressive strength versus time results for both actual values and the predicted values per equation 4 7 for Mix #1. The results of the compressive strength versus time for each mortar mixture can be found in Appendix C. Table 4 3 is a summary of the strength development parameters computed for each mortar mixture The observed data from the compressive strength of cubes versus time compare relatively well with the calculated data per the exponential method. Table 4 3 is a data summary of the curve fit / strength development parameters for each of the mortar mixes. The Ea was determined in the same manner by determining the best fit straight line from the natural logarithm of the val ues versus the inverse of the absolute temperature T. T he negative slope of the line is multiplied by the universal gas constant R (8.3144 J/mol -K) as shown in Figure 4 6.

PAGE 68

68 Figure 4 -4 Compressive strength of mortar cubes vs. time for each temperature Table 4 -3. Strength development p arameters for each c oncrete m ixture per e xponential m ethod Mix Name Temp (C) Su (psi) (hr) R2 Ea (J/mol) Mix 1 (100% Portland Cement) 8 9330 0.692 146.539 0.9980 37401 23 9828 0.692 76.563 0.9998 38 8422 0.692 31.1063 0.9845 Mix 2 (50% Portland 50% Slag) 8 9762 0.475 559.113 0.9983 39932 23 10436 0.657 162.13 0.9999 38 14150 0.526 108.9 0.9972 Mix 3 (65% Portland 35% Slag) 8 6551 0.464 328.254 0.9992 20643 23 10802 0.464 215.159 0.9999 38 10971 0.464 139.866 0.9967 Mix 4 (5030 20 Blend) 8 6307 0.508 348.302 0.9907 21158 23 9395 0.508 197.126 0.9995 38 13783 0.508 143.523 0.9895 The research presented by Schindler (2002) presents a mathematical inconsistency regarding the calculation of apparent activation energy that is not well documented in his work. Step 4 of the procedure for the calculation of activation energy states The Arrhenius plot can now be constructed, by plotting the natural logarithm of the hydration parameters ( ) for each temperature versus the inverse of the absolute curing temperature. (Schindler, 2002 p.207)

PAGE 69

69 However, the par ameters presented by Schindler (2002), are positive values greater than 1. The natural logarithm positive numbers greater than 1 result in positive numbers. However, the k value obtained via the exponential method per equation 4 7 which is developed from t he ln( ), is negative. The reason for the negative value is not documented or resolved in the literature. Regardless of the inconsistency in the presentation of the methodology for the calculation of Ea using the exponential method, the method appears to calcul ate reasonable values. Figure 4 -5 Activation energy calculation for exponential method Activation Energy Determination via Isothermal Calorimetry The determination of Ea of concrete via the use of isothermal calorimetry testing has been documented by several researchers (Ma et.al, 1994, Schindler 2003, 2004, Broda et.al, 2002, Poole, 2007 and Poole et al 2007a). Similar to the methodology behind the determination of apparent activation energy using the rate of strength gain over time, the determination of Ea via the use of isothermal calorimetry utilizes the measured rate of heat generation of cementitious materials at early ages.

PAGE 70

70 Several researcher have used isothermal calorimetry for the determination of Ea in which the rate of hydration (k) was determined from the linear portions of the heat conduction curves as documented by (Ma et, al. 1994) and the use of the modified Arrhenius equation as documented by (Schindler 2003, 2004, Broda et.al, 2002, Poole, 2007 and Poole et. al, 2007a). The use of isothermal calorimetry data analyzed by the Arrhenius equation has gained much favor in recent years. The calculation procedure outlined by Poole, 2007 is as follows: Revisiting equation 4 2 (used to model equivalent age via compressive strength testing), the calculation of Ea using isothermal calorimetry utilizes the same equation. Per equation 4 2 and the Arrhenius theory for rate processes, Ea is assumed to be independent of temperature. t eTtt TTR E rec r 0 273 1 273 1 (42) The hydration of Portland cement is quantified by the degree of hydration ( ) which is a value between 0 and 1, with 0 being equivalent to no hydration having occurred and 1 indicating complete hydration has taken place. Per t his method, the degree of hydration is calculated to be the ratio of the measured heat by the isothermal calorimeter at time t to the total heat available as determined by chemical methods as shown in equation 4 7 (De Schutter and Taerwe, 1996, DAloia, and Chanvillard, 2002 Schindler, A.K. 2002, Poole 2007). uH tH )( (4-7) Where: = degree of hydration H(t) = total heat evolved as measured by the isothermal calorimeter (J/gram) Hu = total heat available (J/gram) The total heat available for Portland cement is determined by equation 4 8. Hcem=500 3S + 260 2S + 866 3A + 420 AF (48)

PAGE 71

71 + 624 3 + 1186 Where: Hcem = total heat available within Portland cement (J/gram) at = 1, p = the mass percentage of the cementitious component A mathematical relationship used to model the degree of hydration development as suggested by a number of researchers (Schindler and Foll i ard, 2005, Poole 2007a) ts uexp (46) Where: = degree of hydration t = test age (hours) u = u ltimate degree of hydration = shape constant = time constant for strength prediction (hours) The following equation can be used to calculate apparent activation energy, Ea (Schindler, 2004, Poole, 2007) R TT Ec ref c ref 11 ln (49) Where: E, Tref, Tc and R are previously defined Apparent activation energy as outlined by Poole, 2007 is calculated as follows. Time and heat evolution data are recorded from the isothermal calorimetry testing of cementitious material at 4 different temperatures: 15, 23, 38 and 49 C (59, 73, 100, 120F) for t his study. Poole (2007) used 5 temperatures. The data are fit to equation 4 6 at each temperature by solving for u, and using a least squares fit for the exponential function u and are presumed independent of (test) temperature, therefore all u and values are equal to each other for a given sample.

PAGE 72

72 ln ( versus 1/Temperature (K) is plotted. Ea is the slope of the best -fit line multiplied by the negative of the universal gas constant R, as defined in equation 49. The parameters presented by Poole (2007) akin to the values presented by Schindler (2002), are positive values greater than 1. Thus, the exact procedure as presented above must be altered to include the negative of ln( ) for the calculation of Ea. Figure 4 7 is a graph of isothe rmal calorimetry data with respect to degree of hydration (in an effort to calculate apparent activation energy) over time, based on the curve fit parameters developed in table 4 4. Figure 4 -6 Isothermal calorimetry data vs. time for four temperatur es The observed data from the isothermal calorimetric data vs. time tests compare relatively well with the calculated data per the exponential method. The Ea was calculated in the same manner by determining the best fit straight line from the natural log arithm of the k values versus the inverse of the absolute temperature. However this technique utilizes four temperature points rather than three (per the compressive strength testing methods). Consistent with the Ea

PAGE 73

73 calculation using compressive strength, the negative slope of the best -fit straight line is multiplied by the universal gas constant R (8.3144 J/mol -K) as shown in Figure 48. Table 4 -4. Strength development p arameters for each c oncrete mixture per i sothermal c alorimetry / e xpoenential method Temp (C) R 2 Ea M ix 1 (100% Portland Cement) 15 0.716 0.943 22.1 0.9768 34235 23 0.716 0.943 19.9 0.9882 38 0.716 0.943 8.0 0.9943 49 0.716 0.943 5.5 0.9938 M ix 2 (50% Portland 50% Slag) 15 0.927 0.542 68.5 0.9677 50400 23 0.927 0.542 47.4 0.9 801 38 0.927 0.542 15.3 0.9580 49 0.927 0.542 7.9 0.9451 M ix 3 (65% Portland 35% Slag) 15 0.932 0.749 27.4 0.9835 32982 23 0.932 0.749 23.0 0.9816 38 0.932 0.749 11.8 0.9717 49 0.932 0.749 6.5 0.9866 M ix 4 (503020 Blend) 15 1.000 0 .523 56.7 0.9915 37330 23 1.000 0.523 36.1 0.9880 38 1.000 0.523 14.9 0.9640 49 1.000 0.523 11.6 0.9951 Figure 4 -7 Apparent activation energy calculation for exponential method

PAGE 74

74 Comparison of Strength Development Parameters Many researchers c onsider the Arrhenius equation to be the most accurate description of the hydration of concrete (Carino 2004, Poole et. al 2007a, Schindler, 2002). As previously discussed, the ASTM 1074 method uses a rational asymptotic model to describe strength developm ent. One limitation of hyperbolic function is that it assumes zero gain in the property under investigation until setting time is reached. Thus, the argument is that the hyperbolic function as outlined in ASTM 1074 does not model the hydration behavior as well as an exponential model. (Poole et.al, 2007). A comparison of each of the methods used to model strength / rate of hydration per the ASTM 1074 method (compressive cube strength) and exponential /Arrhenius method (compressive cube strength) and the e xponential/Arrhenius method using isothermal calorimetry was performed. Table 4 5 Strength d evelopment p arameters for m ixture 1 (100% Portland c ement) per each method of analysis Compressive strength of mortar cubes Method of Analysis Temp (C) Su k T o R 2 Ea Hyperbolic (ASTM 1074) 8 9330 0.004 17.6 0.9950 35642 23 9828 0.008 5.5 0.9998 38 8327 0.019 1.9 0.9944 Temp (C) Su R 2 Ea Exponential constant 8 9330 0.692 146.539 0.9980 37401 23 9828 0.692 76.563 0.9998 38 8422 0.692 31.106 3 0.9845 Exponential constant & Su 8 12771 0.447 274.092 0.9992 30762 23 12771 0.447 127.968 0.9999 38 12771 0.447 77.2207 0.9967 Isothermal Calorimetry Testing Temp (C) u R 2 Ea Exponential constant 15 0.630 0.997 17.3 0.9942 30418 23 0.752 0.997 22.5 0.9938 38 0.698 0.997 7.7 0.9948 49 0.712 0.997 5.7 0.9926 Exponential constant & u 15 0.716 0.943 22.1 0.9768 34235 23 0.716 0.943 19.9 0.9882 38 0.716 0.943 8.0 0.9943 49 0.716 0.943 5.5 0.9938

PAGE 75

75 Table 4 5 provides a summary of the strength development parameters, with R2 values and calculated Ea for each method used. The hyperbolic method per ASTM 1074, Exponential method constant b, constant & u strength testing), and exponential method constant & u are the most widely accepted methods per the available research. The values of the Ea for the Portland cement ( mix 1) are 35642, 37401, and 34235 respectively, indicating that the three methods of analysis are relatively comparable to each other, for mixes contai ning Portland cement. The data indicates that the goodness of fit for each of the curve parameters is relatively strong, as the lowest per the R2 value obtained was 0.976 using the 15C isothermal calorimetry data. However, this measure of prediction e stimation may not be the most accurate description of the curve fitting abilities. A residual analysis may be a more appropriate method of describing the error in the curve fitting. The residual is defined as the error or unexplained variation after fittin g a regression model. The residuals were calculated for each of the techniques used to determine apparent activation energy. Figure 4 -8 Residual error vs. normalized time for hy perbolic curve fitting method (c ompressive s trength)

PAGE 76

76 Figure 4 -9 Residual error vs. normalized time for exp onential curve fitting method (c ompressive s trength) Figures 4 9 and 410 are graphical representations of the residual error vs. normalized time for each compressive strength analysis technique. Despite the commonly accep ted belief the numerous accounts stating the exponential curve fitting method is more appropriate and more accurate for the explanation of hydration, the se two f igures indicate that the hyperbolic method has lower residuals, indicating a better fit than th e exponential method. Figure 4 10. Residual error vs. normalized time for exp onential curve fitting method (i sothermal c alorimetry)

PAGE 77

77 Figure 4 11 is a graphical representation of the residual error vs. normalized time results for the isothermal calorimetr y analysis techniques. The results clearly indicate that there is a need for a higher order model (Ott, 1988) to properly describe the hydration of cement using isothermal calorimetry. Despite the high R2 values calculated for each best fit curve, there is a significant residual error which has not been reported by researchers in the past. Although the curve fitting equations are able to estimate the hydration to a high degree of perceived accuracy, the residuals have not been accounted for. It may be nece ssary to consider using higher order equations for the accurate description of hydration curves. Summary of Results Apparent Activation Energy Three Ea values for each mixture are summarized in Table 4 6. The values for each Ea were calculated in accord ance with the most widely accepted technique as previously discussed therefore, each column in table 4 6 is arranged as follows (Ferraro and Tia, 2009) : The first column is the mix name. The second column reports Ea calculated for each mixture using the hyperbolic method per compressive strength of mortar cubes (ASTM 1074) The third column reports Ea calculated via the exponential method with constant & u per the strength testing of mortar cubes. The third column reports Ea calculated via the exponential method with constant & u per isothermal calorimetry testing. Table 4 -6. Summary of Ea v alues per t esting and analysis m ethod Compressive streng th of mortar cubes Isothermal Calorimetry Mix Name Hyperbolic calculation (ASTM 1074) Exponential calculation Exponential calculation Mix 1 35642 37401 34235 Mix 2 33688 39932 50400 Mix 3 25757 20643 32982 Mix 4 30013 21158 37330 The values obtaine d for the Ea of mix 1 containing only Portland cement have a large degree of variability regardless of the calculation method. However, the Ea values obtained for

PAGE 78

78 the other mix es containing large amounts of supplementary materials vary significantly. Table 47 provides a percentage difference of the Ea results presented in Table 4 6 between the ASTM 1074 h yperbolic calculation and the two exponential calculation methods obtained via compressive strength of mortar cube testing and isothermal calor imetry tes ting, respectively. Table 4 -7. Percentage d ifferences of Ea per t esting and a nalysis m ethod M ix Name Compressive Strength of Mortar Cubes Isothermal Calorimetry Mi x 1 4.94% 3.95% M ix 2 18.53% 49.61% M ix 3 19.85% 28.05% M ix 4 29.50% 24.38% The v alues calculated via the isothermal calorimetry testing for mix es with large amounts of supplementary cementitious materials are consistently larger than those calculated using the compressive strength of mortar cubes (with the exception of the mix containing Portland only). The most plausible reason for the difference is due to the nature of the testing techniques. The acquisition of the data for the compressive strength testing of mortar cubes is performed over a time period of one day to 56 days. Isother mal calorimetry on the other hand, acquires the data used to calculate Ea over a period of one minute to 14 days. It has been stated by several researchers (Poole et. al 2007a; DAloia & Chanvillard 2007; Kada -Benameur, 2000), that the Ea is dependent on degree of hydration of cement, and will tend to be larger at the earlier stages of the cement hydration process which is reactionrate limited as opposed to the later stages of hydration where the reaction is diffusion related. Parrot and Killoh (1984) indicate that the rate of hydration of Portland cement as represented by heat evolution by which the rate was controlled by nucleation and growth. Thus, the Ea as calculated by the acquisition of the heat of hydration via isothermal calorimetry testing is mor e dependent upon the initial rate of the hydration reaction rather than at later ages. Therefore, the calculated Ea per

PAGE 79

79 the isothermal calorimetry testing will tend to be larger for cementitious systems with large replacements of Portland cement in which t he temperature rate of reaction is disproportionally dependent upon temperature. This concept is discussed in greater detail in Chapter Seven. Summary of Findings Chapter 4 The investigation of the maturity and equivalent age parameters for the concrete in this study results in the following: 1. The apparent activation energy is a critical parameter for the determination of equivalent age of concrete. There are several different techniques available for the analysis of the data obtained via compressive stre ngth or isothermal calorimetry. 2. Despite the fact that there is some literature available stating the hyperbolic curve fitting function is inferior to the exponential curve fitting function for the description of hydration -strength parameters of concrete, this research determined the two curve fitting techniques are equally effective (Ferraro and Tia, 2009). 3. The use of the Arrhenius equation for the prediction of earlyage hydration parameters may not be the best available equation as a higher order equati on may be needed to accurately describe early age hydration of cementitious materials. 4. Ea obtained via isothermal calorimetry testing is more likely to provide larger values than Ea obtained by compressive strength testing of mortar cubes.

PAGE 80

80 CHAPTER 5 SE MI -ADIABATIC CALORIMETR Y TESTING Introduction The term s emi adiabatic has gained favor from the concrete industry over the last several decades. The term adiabatic is defined as no flow of energy between the system and the surroundings due to heat, at each point of the process. It is common to model a process as adiabatic because it occurs so rapidly that there is not enough time for appreciable heat transfer to occur (Gater, 1995, p2 11). In the study of thermodynamics, the term semi adiabatic does n ot exist. However, in 1997, the International Union of Laboratories in Construction Materials, Systems and Structures (RILEM) wrote a technical recommendation (TC 119 TCE) which defines a semi adiabatic calorimeter as a calorimeter where the maximum heat losses are less than 100J/(Kh) (RILEM 1997, p451). The purpose of performing semi adiabatic testing of concrete materials is to obtain the adiabatic temperature rise of the concrete at early ages. The concept of semi adiabatic calorimetry has gaine d favor over adiabatic testing of concrete because the equipment needed for adiabatic testing is far more advanced, and requires the controlled supply of heat to the system over time. Semi adiabatic calorimetry is purely passive in nature, as it only requires the monitoring of time, temperature and heat flux for the acquisition of temperature data. The technique has also become more desirable because researchers have had success performing the test in the field (Riding et al, 2007). Research Significance T he purpose of the conducting research using the semi adiabatic calorimetry device is to determine its viability as an appropriate test method for the prediction of heat of hydration and temperature rise in mass concrete. Some researchers state that semi ad iabatic testing is a viable

PAGE 81

81 method for the determination of temperature rise of concrete (Schindler, 2003, Riding et al, 2007, Poole et al, 2007a), while other researchers state that the semi adiabatic calorimetry method does not adequately assess early age heat of hydration (Wadso, 2003, Brown, 1994). This chapter presents the investigation on the use of semi adiabatic calorimetry for the determination of early age heat of hydration of cement/concrete mixtures. Equipment and Procedure A semi adiabatic ca lorimeter consists of an insulated cylinder which contains two thermocouples and a heat -flux sensor. One of the thermocouples is embedded into the concrete specimen in the center of the calorimeter as shown on the right of Figure 5 1, and the other thermoc ouple is located on the on the exterior. The heat flux sensor is embedded within the insula tion of the semi adiabatic calorimeter. Calibration Procedure Before using the semi adiabatic calorimeter, it must be calibrated to account for the heat loss as prescribed by Nordtest (NT Build 480, 1997), which requires a heated water specimen be placed into the drum and allowed to cool. Deionized water is heated to the approximate temperature the concrete is expected to reach within the calorimeter (typically 50 C), and is monitored over time. The heat loss from the calorimeter is calculated as follows (Poole et al, 2007a): d TTA dt dQoutside Inside) ( (5-1) where: = the rmal conductivity of instrument A= area of heat transfer occurs d = thickness of barrier ; Tinside = the temperature on the inner side of the calorimeter Toutside = the temperature on an outer point of the calorimeter

PAGE 82

82 Equation 5 1 can also be mea sured by the heat flux sensor as shown in Figure 52 which illustrates the principle of a heat flux sensor. In the same fashion as the isothermal calorimetry test, the heat flux is obtained by measuring the temperature difference (typically very small) a cross a known thermal barrier and is expressed in terms of a voltage output. The resultant voltage is used as dQ/dt as per equation 5 1 for the expression as a heat flux or heat loss from the semi adiabatic calorimeter (Poole et al, 2007a). Figure 5 -1. Heat flux sensor The rate of heat loss from the semi adiabatic calorimeter is non linear in nature, especially in the time near the commencement of the test. The correction factor in equation (5 2) for the calibration as developed by Poole et. Al (2007) is the most appropriate for the description of heat loss from the calorimeter as shown by Figure 5 3. ))ln((2 1 f f fluxCtCq dt dQ (5 2) Where: Q = Where: qflux = heat flux measured in terms of voltage, Cf1 and Cf2 are correction factors, t = tim e

PAGE 83

83 Figure 5 -2 Calculated losses for the water calibration test It should be noted that the calculated adiabatic temperature slightly oscillates throughout the test duration. As noted by Poole et al (2007), the use of time as the dependent variable in e quation 5 2 may not be the most accurate method for the calibration of the calorimeter because heat loss is dependent on temperature difference rather than time. However, this research investigated several heat loss equations which incorporated temperature difference rather than time, but none were more accurate than the heat loss equation in equation 5 -2. Calculation of Adiabatic Temperature Rise Once the calibration was completed, the loss equation was applied to the semi adiabatic testing results to obtain adiabatic temperature rise. For this research, two methods of the calculation of the fully adiabatic temperature rise (ATR) as calculated from the semi adiabatic test were used and are described as follows: Quadrel IQ d rum test p rocedure One method p rovides a result of the ATR with time as per the data and calculations performed by Quadrel I service and equation 5 3. Quadrel is a testing company which

PAGE 84

84 manufacturers the semi adiabatic calorimeter, and software used for the calorimeter device used in th is study. PCU CmQ ATR (53) Where: Q = the cumulative adiabatic heat of hydration at the eq uivalent maturity age (J/g) Cm = total cement itious content per volume (kg/m3) U = t he unit weight of concrete (kg/m3) CP = specific heat of concrete Mixture (J/kg/C) The total cementitious content of the concrete is typically a design parameter and is inputted into the Quadrel software accordingly. The unit weight of concrete is typically measured at the time of the concrete delivery when the other plastic properties of the concrete are measured and the companion specimens are created (ASTM, 2008d). The value for the specific heat of concrete is typically estimated by the equation 54 (Poole et.al, 2007). However, the Q term for the cumula tive heat of hydration is not provided by the Quadrel software. It is considered proprietary by Digital Site Systems, Inc. (DSS) the makers of Quadrel I -Service. In order to run the software, a Citrix client server is needed to login to the DSS website, af ter which the semi adiabatic heat drum can be programmed. Thus, in order to run any test, a connection with the DSS server must be established before proceeding to raw data acquisition. Once the connection to the server is established, a 6x 12 cylinder i s placed in the semi adiabatic calorimeter and the system monitors the cylinder temperature and heat flux to establish the ATR. ) )1( ( 1wwaaC C ref C pCWCWC WCWC (54) Where: CP = Specific heat of concrete Mixture (J/kg/C) = unit weight of concrete (kg/m3) Wc, Wa, Ww = weight of cement, aggregate and water per unit volume (kg/m3) Cc, Ca, Cw = specific heat of cement, aggregate and water (J/kg/C) Cref = specific heat of hydrated cement = 8.4xTc+339 (J/kg/C)

PAGE 85

85 University of Texas / Auburn University test proc edure The semi adiabatic test procedure has been researched extensively at the University of Texas and Auburn University for the purpose of predicting the temperature rise in concrete (Schindler, 2002 Poole, 2007). The method involves the same calibration and testing procedure as the IQ Drum method. The only difference between the two methods is that the calculation of the ATR is performed by an Excel spreadsheet in which all of the inputs and results can be measured and calculated directly. C r e e e CuhTTR E t tt WHQ 273 1 273 1 exp )( (55) Where: Qh = rate of heat generation (W/m3) Hu = total heat available (J/kg) Wc, = weight of cementitious material content (kg/m3) Tc = average temperature of the concrete Tr = reference temperature (typically 22.8C / 295.8K) E = activation energy, J/mol R = universal gas constant, 8.3144 J/(mol K) = degree of hydration te = eqivalent age (hours) = shape constant = time constant for strength prediction (hours) Revisiting the derivation of the hydration parameters as presented in equations 4 2 47, the heat evolution can be calculated using equation 55 (Poole et. al 2007). The calculation of the ATR of concrete is an iterative procedure using equations 4 2, 4 6, 4 7, 48 and 5 5 which is the same procedure used t o calculate ATR used by Poole et al 2007. Appendix D provides a detailed explanation of the spreadsheet used to calculate ATR per the Texas/Auburn method. Semi Adiabatic Calorimetry Testing Results Semi adiabatic tests were performed on each of the four m ixes per the large block specimens outlined in Chapters 8 and 9. Table 5 1 provides the mixture proportions for each mix

PAGE 86

86 used in this research. Typically, mass concrete mix tures accepted by FDOT for use in construction have contained granulated blast furna ce slag, fly ash or both, in addition to Portland cement. As stated in Chapter 2, the supplementary cementitious materials are used to create a concrete mix ture which has a lower heat of hydration and ultimately a lower temperate rise The mixes were designed to characterize typical bridge substructure elements, which comprise the majority of mass concrete placed in the State of Florida. This research incorporates the investigation of several concrete mix ture designs incorporating the following: Mix 1 100% Portland cement Mix 2 50% Portland cement and 50% granulated blast furnace slag Mix 3 65% Portland cement and 35% class F fly ash Mix 4 50% Portland cement, 30% granulated blast furnace slag, and 20% class F fly ash Table 5 -1. Mixture d esigns fo r concrete s pecimens Material Mix 1 100% Portland Cement (lb/yd3) Mix 2 50% Portland 50% Slag (lb/yd3) Mix 3 65% Portland 35% fly ash (lb/yd3) Mix 4 5030 20 blend (lb/yd3) Cement 681 341 443 341 Fly ash 0 0 238 136 GGBF Slag 0 341 0 204 W ater 341 341 341 341 Fine Agg 1095 1088 1036 1050 Coarse A gg 1650 1668 1660 1650 A comparison the calculated ATR results for both the Texas / Auburn method of analysis and the IQ Drum method of analysis from Mix #1 is presented in Figure 5 4. The Tex as / Auburn method utilizes the True Adiabatic curve as the ATR, where the IQ Drum curve is denotes the ATR for the IQ drum technique. The measured and calculated curves are the observed temperature readings prior to analysis.

PAGE 87

87 Figure 5 -3. Adiaba t ic temperature rise for Mix 1 The results indicate that after 72 hours of age, Mixes are within 1.1% of each other. However, the IQ drum does indicate a slightly larger ATR after 60 hours. Regardless, the two techniques results indicate that the methods a re comparable according to precision statements made by (Moritabo, 1998 and Poole et.al, 2007). Prior to the development of the Texas / Auburn method of analysis, the large -scale block specimens were created and the semi adiabatic calorimetry testing was performed on each of the mixes delivered by the concrete supplier. The initial temperatures for each mix are listed in Table 52, which indicate that the delivery temperature of each mixture was higher than that of typical laboratory temperatures of 23 C Table 5 -2. Delivery t emperatures for c oncrete m ixes Mix Name Delivery Temperature (C) Mix #1 35.0 Mix #2 32.2 Mix #3 30.6 Mix #4 28.3

PAGE 88

88 Since temperature and heat flux logging equipment was not made available at the time of the large -scale block sp ecimen creation, the IQ drum method was used for the analysis. However, at the time of delivery of mix 1, the FDOT computer server experienced a communication failure with the DSS server and initiation of the semi adiabatic test per the IQ drum was delayed approximately 7 hours. The curves shown in Figure 5 3 indicate the discrepancy of the ATR of the calorimetry specimens created in the lab (after the experiment) and the specimen created with the concrete delivered to the laboratory with large scale block specimen. The curves indicate that the absence of the first 7 hours of data logging as per the delivered mix, will result in a curve that does not represent the true nature of the ATR of concrete as indicated by the lower calculated ATR. Literature indicates that concrete mixes with higher mixing temperatures will result in higher early age ATR (ACI 207.1, 2005). Subsequent to the testing of Mix 1 delivered to the FDOT, the mixture was recreated in the laboratory at a temperature of 23 C in an effort to get the ATR per the monitoring of the specimen for the entire age of the specimen. The specimens delivered for the large -scale testing blocks experienced approximately an hour of delay due to transportation from the concrete mixing plant. The ATR was calculat ed via both methods as shown in Figure 5 4. The ATR for the concrete specimen created from the delivered concrete (mix 2) and the mix recreated in the laboratory (mix 2 ) are shown in Figure 5 5. The ATR calculated for Mix 2 indicates ATR difference of app roximately 18C between the delivered mix verses the mix created in the lab, despite the fact that the mixes had a respective mixing temperature differential of 9 C ACI 207.1 indicates that for a Type I Portland cement, the ATR difference between the two mixes should be approximately 4.5 C This

PAGE 89

89 indicates that the ATR concrete created with blast furnace slag does not behave the same manner as predicted by historical curves for concrete made with ordinary Portland cement. Figure 5 -4. Adiabatic temperatu re rise curves for Mix 1, delivered mix and lab m ix Figure 5 -5. Adiabatic temperature rise for Mix 2, delivered m ix and lab m ix The ATR for each of the mixes is presented in Figure 5 6, where each ATR was calculated from the delivered mix in conjunction with the large scale specimens with the exception of Mix

PAGE 90

90 1. The results indicate that Mix 1, which does not contain any supplementary cementitious materials, has a significantly higher ATR than the mixes with supplementary cementitious materials, (other th an Mix #2) despite the fact the mixing temperature (only for Mix 1, mixed in the lab) was lower than that of the other three mixes. Mix 2 achieved a larger ATR than each of the other mixes. However, this result could have been caused by the fact that Mi x 2 had the highest delivery temperature of the mixes (because Mix 1 was made in the lab). However, this is most likely not the case, since the large scale block specimen for Mix 2 achieved the largest core temperature, per the data presented in Chapter 9. Furthermore, the compressive strength of mortar data presented in table 4 2, indicates that Mix 2, cured at 38 C obtained the highest limiting strength of any of the mixes. Therefore, the physical data indicates that the replacement of Portland cement wi th blast furnace slag will result in continued hydration at later ages. Comparatively, the mixes without the addition of slag (Mix 1 and Mix 3) have higher rates of hydration at earlier ages, based on the ATR. Figure 5 -6. Adiabatic temperature rise for each mix

PAGE 91

91 The data obtained from Mixes 3 and 4 indicates that the replacement of Portland cement with fly ash will result in a lower ATR as well as a lower rate of hydration approximately 12 hours after mixing. This result is consistent with literature (Tay lor 1997, Gajda, 2007). Interestingly enough, Mix 4 (similar to Mix 2) demonstrates a lower rate of ATR at early ages; yet, after approximately 12 hours, the rate of temperature increase is comparable or larger to that of Mix 3 afterward. The ATR of Mix 4 becomes higher than Mix three at approximately 125 hours of age. One potential reason for the delay in reactivity or temperature rise of the Mixes containing blast furnace slag is thought to be the impervious coatings of silica and alumina that form on slag particles early in the hydration process. The presence of calcium hydroxide (forming from the hydration of Portland cement) is a potential reason for the activation of the slag particles which will potentially continue to hydrate after the initial activ ation (Mindess et.al, 2003). Summary of Findings The investigation of semi adiabatic testing for the concrete in this study resulted in the following : The calculation of adiabatic temperature rise of concrete can be performed using semi adiabatic calorim etry. The replacement of Portland cement with supplementary cementitious materials will result in lower early age temperatures (48 hours or less), and lower ultimate temperatures. After 48 hours, the reaction rate of concrete containing large replaceme nts of blast furnace slag when combined with Portland cement will result in higher rates of temperature rise when compared with concrete made with Portland cement alone, or Portland cement with class F fly ash.

PAGE 92

92 CHAPTER 6 SURE CURE/ADIABATIC CALORIMETRY T ESTING The Sure Cure device was designed to be used in concrete prestressed / precast facilities for the simulation of high temperature curing of concrete cylinders used to make earlyage comparative determinations regarding formwork removal for larger ste am cured structural elements. It can also be used to determine temperature rise of concrete at early ages and accordingly, it was used in this experiment for the simulation of adiabatic conditions for concrete specimens. Research Significance The purpose of the conducting research using the Sure Cure adiabatic calorimetry device is to determine its viability as a potential test method for the prediction of heat of hydration and temperature rise in mass concrete. No published research is available regardin g the use of the Sure Cure device for adiabatic testing. Since there is research stating that the semi adiabatic calorimetry method does not adequately assess early age heat of hydration (Wadso, 2003, Brown, 1994), this Chapter investigates the potential use of alternative calorimetry method for the measurement of temperature rise in concrete. The isothermal calorimetry testing method has shown promising results for the measurement of heat of hydration and temperature rise in cement. However due to the sma ll specimen size which can be accommodated by the isothermal calorimeter (30g or less), it cannot be used on concrete specimens. Therefore, an alternative or supplementary method could prove to be beneficial for the prediction of heat of hydration and temp erature rise of concrete. Per RILEM TC 119 TCE, a calorimeter which can limit a concrete specimens temperature loss to less than 0.02K/hour or less, may be used for the calculation of ATR. Referring to Gater (1995), it is common to model a process as adiabatic because it occurs so

PAGE 93

93 rapidly that there is not enough time for appreciable heat transfer to occur (Gater, 1995p211). Thus, similar to the semi adiabatic test method in Chapter 5, in the strictest of definitions the term adiabatic testing of the temperature rise of concrete is not physically possible. However, per the adiabatic definition per RILEM TC 119, similar to the research performed by Ng et.al (2008), the loss of temperature per this method is prevented by the introduction of a co ntrolled quantity of heat into the system. This Chapter discusses the viability of using the Sure Cure device for measurement of temperature and heat rise of concrete Equipment The Sure Cure system is designed to control the temperature of concrete specim ens in several ways. The system can receive temperature from an outside source such as a concrete specimen, or it can be inputted into the computer controller to provide a user defined time temperature history. Figure 6 -1. Schematic of S ure Cure s yste m with h ydration c hamber The sure cure system used for the adiabatic testing in this research is essentially composed of three major components as illustrated by Figure 6 1. The computer operates the software which provides signals to the controller box. T he controller box sends and receives temperature signals to and from the hydration chamber. The hydration chamber contains insulation and

PAGE 94

94 heating elements used to physically control the temperature. Additionally, the hydration chamber contains two thermoco uples, one in the center of the chamber acquires the temperature at the center of the concrete mass, the other monitors the temperature within the center of the insulation. The temperature within the insulation of the hydration is controlled by the tempera ture of the concrete itself. The recorded timetemperature curve of the concrete is the natural hydration curve. Figure 6 2 shows the orange hydration chamber and the micro computer controller. Figure 6 -2. Photograph of the Sure Cure system with the c omputer, hydration chamber and controller box Test Procedure Since the Sure Cure System has never been used to establish the relationship between the temperature rise of concrete, it was necessary to perform a series of trial experiments to establish the p otential of using the sure cure system for the determination adiabatic temperature rise of concrete. As per RILEM TC 119, From the point of view of the reliability in determining the adiabatic temperature increase the temperature of the sample in an adia batic calorimeter is at

PAGE 95

95 any time very close to the temperature of the adiabatic temperature so the influence of the change in the reactivity of the cement with temperature is taken into account (RILEM 1997 p.317). The Sure Cure system can be programmed so that the hydration chamber temperature is the same as, or offset from, the temperature of concrete within. As such, several trials of experiments were performed prior to the large scale block experiments to establish a temperature configuration in whi ch the concrete would experience a true adiabatic temperature rise, rather than gaining temperature at a faster or slower rate, due to the heat input as provided by the hydration chamber. The experiments were designed to calibrate the temperature rise of t he concrete within the Sure Cure hydration chamber to match the temperature rise within the semi adiabatic calorimeter. Accordingly, the semi adiabatic temperature of a 6x12 concrete specimen was obtained per the procedure outlined in Chapter 5. Afterwar d, the following procedure was developed for acquisition of the adiabatic temperature rise for the Sure Cure system: A 25 lb sample of fresh concrete was taken from each mix and placed in the hydration chamber. A total of four hydration chambers (individual specimens) can be monitored for a single concrete mix The concrete obtained for the Sure Cure and semi adiabatic device was taken from the same concrete mixture. The initial calibration trial for the Sure Cure testing involved the programming of the sur e cure system to three offset temperatures. An offset temperature is defined as the temperature of the concrete as recorded by the internal thermocouple, plus the temperature of the hydration chamber. Thus, an offset of 1.0F indicates that the temperatur e within the hydration chamber is one degree Fahrenheit less than the concrete temperature (the offset temperature is presented in Fahrenheit in this chapter in an effort to maintain consistency with the Sure Cure program).

PAGE 96

96 Sure Cure Calorimetry Testing R esults The calibration involved programming the hydration chamber to offset temperatures of 0.0F, 0.5F, and 1.0F. The results from the initial calibration are presented in Figure 6 3, where the offset chambers are presented in the legend and the semi adiabatic result is denoted as Semi -Adiabatic. The data indicates that the temperature offsets of 0.0F, 0.5F rise to a limit of approximately 92C and thereafter become constant. The limit was set in an effort to prevent damage to the hydration chamb ers. The hydration chamber programmed with a 1.0F offset achieved a ATR curing history lower than that obtained by the semi adiabatic testing device. Figure 6 3. Sure Cure and s emi adiabatic temperature vs. time Per the initial calibration testing, it was decided to set the hydration chambers to an offset temperature between 0.5F, and 1.0F to better replicate the time -temperature history of the semi adiabatic test. Therefore the calibration procedure was repeated with offset temperatures of 0.5F 0.6F, 0.7F,and 0.8F. The results presented in Figure 6 4 indicate that the time -

PAGE 97

97 temperature curves between an offset temperature of 0.7F and the semi adiabatic test are relatively comparable. Figure 6 4. Sure Cure and s emi adiabatic temperatur e vs. time As a result of the calibration testing, it was decided to use an offset temperature of 0.7F for the Sure Cure testing for the large -scale block experiments and the testing results are presented in Figure 6 5. The Sure Cure adiabatic test resul ts indicate the test method does replicate the ATR of concrete in a similar fashion as the semi adiabatic test. A thorough comparison of each of the calorimetry testing methods is presented in Chapter 7. Figure 6 5. Sure Cure temperature rise for large scale block experiments

PAGE 98

98 Limitations of the Test Method The Sure Cure testing system has several potential limitations which may not permit its use for adiabatic testing of concrete. The Sure Cure computer programming is limited to a minimum temperature o ffset of 0.1F. Per Figure 6 4, the offset temperature of 0.7F provides a comparable timetemperature history as the semi adiabatic temperature testing. However, research indicates that the semi adiabatic test does not replicate the true ATR of concrete (Wadso, 2003). Additionally, Sure Cure test performed with an offset temperature of 0.6F resulted in a ATR of 13C, approximately 20 percent higher than the concrete with an offset temperature of 0.7F/semi adiabatic test. Results presented in Chapter 7 indicate that the energy rise of cementitious materials tested via the semi adiabatic calorimetry test is approximately 10% lower than the energy rise of cementitious materials tested via isothermal calorimetry testing. However, the smallest allowable i ncrease in offset temperature via the sure cure test is 0.1F. Thus, the precision limitations of the Sure Cure system, do not easily resolve the differences in the ATR produced for each offset temperature. The concrete composed of Portland cement and ce ment with fly ash tested via the Sure Cure method showed a decrease in temperature after 100 hours of testing. This temperature decrease may indicate a lack of necessary temperature or energy input with a hydration chamber offset of 0.7F. Since a temper ature offset between 0.6F and 0.7F cannot be programmed, this method may not be suitable for the accurate calculation of ATR. Summary of Findings The investigation of Sure Cure adiabatic testing for the concrete in this study resulted in the following : The calculation of adiabatic temperature rise of concrete may be performed using Sure Cure Adiabatic calorimetry.

PAGE 99

99 The replacement of Portland cement with supplementary cementitious materials will result in lower early age temperatures (48 hours or less), and lower ultimate temperatures which is similar to the results obtained via the semi adiabatic testing method. The Sure Cure testing device may not be suitable for the accurate calculation of ATR due to the limitations in the allowable temperature off sets. Further is testing is needed to establish the potential of the Sure Cure test method to be used for ATR

PAGE 100

100 CHAPTER 7 COMPARISON OF CALORIMETRY TESTING METHODS Introduction The purpose of evaluating the three calorimetry testing methods in this researc h was to determine the potential for each to serve as the method to provide the primary input parameter for the modeling of the energy rise of cementitious materials and, ultimately, mass concrete. This chapter discusses the comparison of the test results provided by each calorimetry testing method and the potential implications for the use of each method as a predictor of physical properties of cementitious materials. Research Significance There is currently insufficient research which compares adiabatic and semi adiabatic testing of concrete (Ng.et.al, 2009). Although RILEM TCE 119 (1998), and Morabito (1998) address the specifications required to construct both types of equipment, there is presently no known, published research available which compares c ommercially available equipment using either of the two methods. The comparison of test data between the Sure Cure adiabatic test and the Quadrel IQ Drum semi adiabatic test has not yet been performed. There is also little research available (Wadso, 2003) which compares the energy/temperature rise of cementitious materials achieved between the semi adiabatic calorimeter and the isothermal calorimeter. This chapter, then, fills this research void by presenting the results of mortar specimens tested using bo th test methods. The direct comparison between isothermal calorimetry testing and semi adiabatic testing of mortar has not yet been performed. The standard test method for the heat of hydration of hydraulic cement per ASTM C186 was performed on the cement itious materials for each concrete Mixture. The heat of hydration

PAGE 101

101 test is the only standardized test method (per ASTM) and, therefore, it was decided to compare the data obtained from the standardized test method and the other calorimetry tests performed f or this research. Calorimetry Testing Results Comparison of Sure Cure adiabatic calorimeter and semi -adiabatic calorimeter The comparison of test data between the Sure Cure adiabatic calorimeter (hydration chamber) and the Quadrel IQ Drum semi adiabatic c alorimeter was performed for each large scale mix as per the procedures provided in Chapters 5 and 6. The results shown in Figures 7 1 and 7 2 indicate that concrete tested within the Sure Cure hydration chamber did not achieve the same temperature increas e as compared to that contained within the IQ Drum semi adiabatic calorimeter. Figure 7 1. Comparison of time temperature histories between Sure Cure system and the IQ drum for Mix 1 As discussed in Chapter 6, the Sure Cure system did not provide res ults which were precise and accurate enough to be considered for the use in the calculation of ATR. Therefore, the Sure Cure hydration chamber in its current configuration is not recommended for the

PAGE 102

102 measurement of ATR of concrete. However, the incorporation of several small software and hardware changes designed to increase the accuracy of the system which follows the recommendations of RILEM TCE 119 or Ng et. al (1998) could make the Sure Cure a viable testing technique. More research is accordingly needed to determine if the Sure Cure system is viable for adiabatic testing. Figure 7 2. Comparison of time temperature histories between Sure Cure system and the IQ drum for Mix 2 Comparison of semi adiabatic calorimeter and the isothermal calorimeter The en ergy rise for both the semi adiabatic calorimeter and the isothermal calorimeter was calculated for each test. As discussed in Chapter 3 and Appendix A, the isothermal calorimetry test method measures power directly (from voltage measurements), and the int egration of the power as shown in Equation A 4 is used to calculate the energy rise (per mass). The energy rise (per mass) using the semi adiabatic calorimeter is determined by measuring the adiabatic temperature rise, and, by applying the first law of t hermodynamics, the principle of energy conservation, to convert temperature rise to energy rise per Equation 71. TCmQp (71)

PAGE 103

103 Where: Q = the change in energy or energy rise (J) m = mass of concrete (g) Cp = Specific Heat Capacity (J/gK) = the change in energy or energy rise (K) As discussed in Chapter 5, the assumption that an adiabatic condition exists while using the semi -a diabatic calorimeter is not a reliable one, since a known amount of heat is lost (and measured) per the test method. Therefore, the assumption that the first law of thermodynamics applies to the conversion of temperature rise to energy rise is not necessar ily valid for the conversion from adiabatic temperature rise to energy rise of concrete. Comparison of cement and concrete The isothermal conduction calorimeter is designed for use on relatively small specimens (420g) and the semi adiabatic calorimeter is primarily designed for use on specimens approximately three orders of magnitude larger. The calculated energy with respect to equivalent age for the isothermal test method as well as the semi adiabatic test method is presented in Figure 7 3 for Mix 1. T he isothermal test was performed at temperatures of 15C, 23C, 38C and 49C. The semi adiabatic test was performed with an initial temperature of 23C (laboratory conditions). The results indicate that the energy rise with respect to equivalent age is su bstantially different with regard to each testing method. The semi adiabatic test obtained a maximum energy of (241J/g) at 250 hours where energy measured for each of the isothermal tests was above 305J/g. Thus, the calculated energy rise for the cementiti ous materials is relatively larger than that measured by the isothermal calorimetry test.

PAGE 104

104 Figure 7 3. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 1 Referring back to Figu res 3 4 and 3 6, the energy increase with respect to time is significantly larger at larger temperatures. However, Figure 7 3 indicates that the values for energy increase for the isothermal calorimetry with respect to equivalent age is virtually equal re gardless of the test temperature. However, the energy rise at lower temperatures is slightly larger with respect to equivalent age per Peng et. al (2009), the testing resulted in similar findings regarding energy rise and equivalent age for isothermal cal orimetry testing. These findings are likely to be attributable to the hydration kinetics of Portland cement at higher temperatures which are similar to the phenomenon discussed in Chapter 4, where the limiting strength decreases as the curing temperature increases (Tank, 1988; Carino, 2004). At higher temperatures, the hydration (strength evolution) of concrete takes place at a faster rate. However the hydration at later ages (ultimate strength) will be reduced. This crossover effect as described by V erbeck and Helmuth (1968) and Carino (2004), is the concept that the hydration

PAGE 105

105 products of Portland cement which have experienced high temperature do not become uniformly distributed into the hardened cement paste matrix due to rapid product formation at e arly ages. Escalante Garcia and Sharp (2001) attribute the crossover -effect to an observed increase in porosity for cements which are cured at higher temperatures. Figure 7 4. Comparison of equivalent age energy histories between semi adiabatic calori metry system and isot hermal calorimetry for Mix 2 The calculated energy with respect to equivalent age for the isothermal test method as well as the semi adiabatic test method is presented in Figure 7 4 for Mixture 2. The results indicate that the energy rise with respect to equivalent age is substantially different with regard to each testing method. The semi adiabatic test showed a maximum energy of (210J/g) at 250 hours whereas energy measured for each of the isothermal tests was above 325J/g (at temper atures of 38C and 49C). While similar to the results presented in Figure 7 3, the calculated energy rise for the cementitious materials presented in Figure 7 4 is relatively larger than when measured by the isothermal calorimetry test. However, unlike the results obtained for Mixture 1, the results of the testing of energy rise for the isothermal calorimetry for Mixture 2 (with 50% slag replacement) indicate dissimilar

PAGE 106

106 values for each temperature with respect to equivalent age. The energy rise at higher temperatures is larger with respect to equivalent age from testing. Figure 7 5. Mortar cube compressive strength vs. time for Mix 2 There is insufficient research as to why this phenomenon occurs. However, the strength testing results presented in Ta ble 4 2 indicate that the ultimate/limiting strength is higher for the specimens cured at higher temperatures for Mix 2. Therefore, the crossover effect is not observed with regard to cements with large replacement of blast -furnace slag. Similar results are observed with regard to compressive strength testing vs. age per the results presented in Figure 75 for isothermally cured cube specimens used calculate Ea. Chapter eight reports similar results with respect to strength versus equivalent age test resu lts. Additionally, similar results are reported by Schindler (2003) and Brooks et. al (2007) for compressive strength verses time data. All of the results indicate that the crossover effect does not occur when large amounts of granulated blast furnace sl ag is incorporated into a mixture to replace Portland cement. To date, there has not been a reported explanation as to why this phenomenon occurs. One potential reason may be due to the formation of different reactants at higher temperatures in Portland

PAGE 107

107 cement with large replacements of slag. Another possible reason for the phenomenon is that the hydration of a blend of Portland and slag is not diffusion -controlled at later ages. In order to ascertain the reason why Portland cement replaced with large porti ons of slag does not experience the crossover effect at later age more research is needed. Figure 7 6. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 3 Figure 7 7. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 4

PAGE 108

108 The calculated energy with respect to equivalent age for the isothermal test method as well as the semi adiabatic test method is prese nted in Figures 7 6 and 7 7 for both Mixtures 3 and 4. Both sets of results are consistent with the data in Figures 7 3 and 7 4 where the energy rise with respect to equivalent age is substantially different between the semi adiabatic test and the isothermal calorimetry test. Mixture 3 (which contains 35% fly ash) behaves in a manner similar to the Mixture containing ordinary Portland cement in that the energy rise is larger at lower temperatures for a given equivalent age. Mixture 4, however, behaves more like Mixture 2, lacking a demonstration of the crossover effect per the energy rise curves. This behavior is most likely due to the presence of blast furnace slag (20%) within the Mix. Mortar testing As stated earlier, t he isothermal conduction calorim eter is designed for use on relatively small specimens (4 20g) and the semi adiabatic calorimeter is primarily designed for use on specimens approximately three orders of magnitude larger. However, it is possible to create mortar mixes which can be tested in each device. Thus, the best method of directly comparing the results from both test methods is to obtain the energy rise data as measured by each device. Mortar specimens consisting of cementitious materials used to create Mix 1 and Mix 3 were then created and tested in each calorimeter accordingly. Mixes 2 and 4 were not created for this portion of the experiment since this portion of the testing was not anticipated and there was not enough cementitious material available from those mixes to create specimens for semi adiabatic testing. The isothermal test was performed on mortar specimens at temperatures of 23C, 38C, and 49C. The semi adiabatic test was performed on mortar specimens with an initial temperature of 23C (laboratory conditions).

PAGE 109

109 Figu re 7 8. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and the IQ drum for mortar Mix 1 Figure 7 9. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and the IQ drum for m ortar Mix 3 Figures 7 8 and 79 indicate that the isothermal specimens cured at higher temperatures experienced the crossover -effect similar to the ordinary Portland cement Mixture (without sand)

PAGE 110

110 in Figure 7 3. The energy rise obtained by the semi adiabati c calorimeter is lower than the energy rise measured by the isothermal calorimeter. However, unlike the data presented in Figures 7 3, 7 4, 7 6, and 7 7 the data presented in Figures 7 8 and 79, the material is the same exact mortar Mix. Thus, potential d iscrepancies in the energy difference between isothermal calorimetry and semi adiabatic calorimetry are not due to differences in material behavior between cement paste and concrete. These discrepancies are most likely due to the differences within the tes ting equipment. The isothermal calorimeter was designed to be a device to measure heat flux with a high degree of precision and accuracy. The isothermal calorimeter was designed such that the entire heat from the cement sample passes across the heat flux sensor to a heat sink on the other side as illustrated by figure 7 10. Thus, the heat flux is one directional. The internal heat flux sensors, presence of heat sinks, several layers of insulation a reference specimen (which produces no heat) and the calibration technique itself all combine to facilitate the accurate, reliable and repeatable test results. Figure 7 10. Cross section of isothermal calorimetry device

PAGE 111

111 The semi adiabatic calorimeter was designed to obtain the temperature rise of the concrete in the field. Thus, the semi adiabatic device was never intended to serve as a high precision instrument. However, it has been used with a satisfactory degree of success for the prediction of temperature rise in concrete structures (Riding et. al, 2007). O ne potential reason for the difference in the energy rise values between the isothermal calorimetry test and the semiadiabatic test may be attributed to the physical location of the flux sensor in the semi adiabatic calorimeter. As illustrated in Figur e 7 11, the flux sensor does not make direct contact with the concrete specimen. Although the configuration for the isothermal calorimeter is somewhat similar (as illustrated by Figures in the A 4 and A -4 in Appendix A), the primary difference between the two devices is that the isothermal calorimeter is equipped with a heat sink where the semi adiabatic calorimeter is not. Therefore, it is plausible that a portion of the heat flux not measured by the semi adiabatic calorimeter is measured by the isothermal calorimeter. The illustration in figure 7 11 illustrated a loss of heat on all sides of the concrete specimen, not just the side exposed to the flux sensor. Therefore, it is possible that the heat loss is not measured, as there is potential for varying de grees of heat loss on each side. Figures 7 3, 7 4, 7 6, 77, and 7 8 illustrate that the slopes shown on the energy vs. equivalent age graphs are larger for the isothermal calorimetry test compared to the semi adiabatic test after the initial hydration cu rve (approximately 50 hours). Thus, the difference discrepancy in the energy curves could be attributed to the inability of the semi adiabatic calorimeter to accurately and precisely measure lower quantities of heat loss or heat flux evolved by Portland ce ment based materials at later ages.

PAGE 112

112 Figure 7 -11. Cross section of semi adiabatic device Comparison of solution calorimetry to isothermal and semi -adiabatic calorimetry The only standardized method currently used to determine the heat of hydration of h ydraulic cement is the solution calorimetry method as per ASTM C186. Since a portion of this research was dedicated to the comparison of calorimetry methods, it was decided that a baseline should be established which was incorporated into the standardized test method. The testing was performed at the Construction Technology Laboratories Inc. (CTL) facility in Skokie, Illinois. Prior to the shipment of the materials, appropriate proportions of the cementitious materials were combined at the FDOT SMO. Table 71 provides a summary of calorimetry values for each testing method. Isothermal calorimetry data is reported for energy rise at 7 days of actual time and equivalent age for each temperature. The calculated energy rise per the semi adiabatic calorimetry t esting is reported at 7 days of actual time and equivalent age. Solution calorimetry results are likewise reported.

PAGE 113

113 Table 7 -1. Comparison of energy rise values for each calorimetry method for each m ix Mix 1 Mix 2 Mix 3 Mix 4 Energy (J/g) Energy (J/g) Energy (J/g) Energy (J/g) Temperature C Isothermal Calorimetry at 7 days (Time) 15 287 225 214 213 23 307 274 229 245 38 343 347 278 297 49 N/A N/A N/A N/A Temperature C Isothermal Calorimetry at 7 days (Equivalent Age) 15 323 258 235 241 23 307 274 229 245 38 295 288 222 256 49 289 299 221 276 Semi -Adiabatic Calorimetry at 7 days (Time) 252 255 194 200 Semi -Adiabatic Calorimetry at 7 days (Equivalent age) 227 236 153 151 Solution Calorimetry at 7 days 297 252 217 258 Of the various c alorim etry test methods uses in this study, the isothermal calorimetry testing at a temperature of 23C and 7 days resulted in an energy increase most comparable to the solution calorimetry for each of the testing methods and mixes performed. With the exce ption of the ternary blended m ix (Mix 4) the energy measured obtained by isothermal calorimetry was slightly larger than the energy obtained by the solution calorimetry. This could be attributed to the fact that the single operator precision of the solutio n calorimetry (as reported by ASTM C186) has been found to be 14.8 J/g, indicating a lack of precision of the solution calorimetry method. However, since the measured energy rise values obtained by the isothermal calorimetry measurements are larger than th ose obtained by solution calorimetry, it may be possible for the isothermal calorimetry method to measure heat of hydrations whereas the solution calorimetry method cannot. These findings are consistent with those of Wadso (2003).

PAGE 114

114 The semi adiabatic test ing resulted in an underestimation of the energy rise for each of the mixes with the exception of Mix 2 which incorporated 50% slag replacement. This therefore confirms the suggestion that the use of the semi adiabatic method does not measure all of the en ergy evolved from a cementitious system. Summary of Findings The comparison of the different calorimetry methods resulted in the following: The Sure Cure system does not have the necessary capabilities to accurately provide heat to concrete to represent adiabatic conditions as per RILEM TCE 119. Therefore, the use of the Sure Cure hydration chamber is not recommended for measuring the energy rise in concrete. The semi adiabatic test method does not measure as much of the energy evolved from a cementitiou s system when compar ed with isothermal calorimetry and solution calorimetry. The Isothermal conduction calorimetry method allows the measurement of the largest amount of energy evolved from a cementitious system. Accordingly, the isothermal calorimetry met hod is the most conservative method for measuring the energy rise or heat of hydration of cements. The total energy rise measured in Portland cement systems at cured higher temperatures is lower than those systems cured at lower temperatures. Accordingly, the cross over effect, the phenomenon in which Portland cements cured at higher temperatures will show a decreased ultimate strength at later ages is observed via the energy rise in concrete as measured by isothermal calorimetry. The cross over effect is not present in cementitious systems which have large replacement of Portland cement with granulated blast furnace slag. The solution calorimetry method is acceptable for measuring the heat of hydration of a system. However, it is limited because it o nly allows for a single point of measurement for each sample. Isothermal calorimetry measures heat of hydration over time. Thus, the use of isothermal conduction calorimetry is preferred for the potential modeling of the hydration process of cementitious m aterials. The findings herein are similar to those by Wadso (2003) who concluded that the only acceptable method for the calculation of temperature rise in concrete is the use of isothermal calorimetry (when compared with semi adiabatic and solution calo rimetry).

PAGE 115

115 CHAPTER 8 LARGE -SCALE BLOCK EXPERIME NT PHYSICAL TESTING The testing of the early age concrete subsequent to the creation of the large scale block specimens was performed using physical laboratory testing methods. The results obtained from te sting the large block specimens will serve as potential input parameters the prediction of the behavior of mass concrete. Mixture Design M ass concrete mix tures that are accepted by FDOT for use in construction have t ypically contain granulated blast furnace slag, fly ash or both, in addition to Portland cement. The purpose for using the supplementary cementitious materials is to create a concrete Mix ture which has a lower heat of hydration and ultimately a lower temperate rise. This research incorporat es the investigation of several concrete Mix ture designs composed of the following: Mix 1 100% Portland cement Mix 2 50% Portland cement and 50% granulated blast furnace slag Mix 3 65% Portland cement and 35% class F fly ash Mix 4 50% Portland ceme nt, 35% granulated blast furnace slag, and 20% class F fly ash Table 8 -1. Mixture d esigns for large -s cale b lock s pecimens Material Mix 1 100% Portland Cement (lb/yd3) Mix 2 50% Portland 50% Slag (lb/yd3) Mix 3 65% Portland 35% fly ash (lb/yd 3 ) Mix 4 5030 20 blend (lb/yd 3 ) Cement 681 341 443 341 Fly ash 0 0 238 136 GGBF Slag 0 341 0 204 Water 341 341 341 341 Fine Agg 1095 1088 1036 1050 Coarse A gg 1650 1668 1660 1650 Th e mix designs for the large concrete blocks shown in Table 81 we re designed with 0.50 water to cementitious material ratio in an effort to be compatible with isothermal calorimetry testing. The coarse and fine aggregates were adjusted to reflect the volumetric differences due to the density differences in each cementit ious material. Cement has a specific gravity of

PAGE 116

116 approximately 3.15 where fly ash and slag are approximately 2.5 and 2.9, respectively. The approximate size of each large -scale specimen is 3.5 x 3.5 x 3.5, with an approximate weight of 5,800 lbs. Figure 71 is a photograph of the typical large block testing setup. Figure 8-1. Large s cale b lock t esting c onfiguration Concrete Specimen Creation and Curing The concrete used for the large -scale and small -scale specimens was not batched at University of Florida or FDOT laboratories since approximately 100ft3 was required for each Mixture. The maximum volume for a concrete mix the FDOT State Materials Office can create is approximately 16ft3. As a result, the concrete was purchased from Florida Rock Industrie s ready -Mix plant in Gainesville, Florida. At the time of delivery of the concrete for the large -scale block specimens, companion cylinder specimens were also created for the physical testing. The large -scale and companion specimens were cast in accordance with FDOT Standard specification 346, with the exception of 3463, as the slump was higher than three inches (typically 8). Two sets of 8 high 4 diameter cylindrical specimens were created from the delivered concrete. One set of the cylinder specimens was placed in a curing environment per ASTM C

PAGE 117

117 192 and ASTM C 511. As second set of specimens was match cured with the temperature history of the large -scale block (which remained in the laboratory), using a Sure Cure system (discussed in Chapter 6)manufactured by Products Engineering, Inc. In an effort to create a temperature history within the match cure cylinders identical to a single point within the large scale block specimen, the thermocouples from the Sure Cure system were positioned at the midpoint between two corners along an edge of the non -insulated block specimen as shown in Figure 8 2. Figure 8-2. Thermocouple location for Sure Cure s ystem Thermocouples were located at the midpoint of the longitudinal axis of the block two inches below the concrete surface and two inches from the side wall of the formwork. This location was chosen because it would provide an area of relatively low temperature history, therefore providing a relative equivalent age lower than that of the inner core of the bloc k. Maturity and Equivalent Age Concrete with lower temperature history will result in lower equivalent age, maturity, and relative strength despite the fact that the actual age is the same (Tank, 1988; Carino, 2004). Therefore, it was decided to monitor a portion of the block near the surface which would result in

PAGE 118

118 relatively low strength and relatively high stresses as discussed in Chapter Two. The cylindrical specimens were placed in the laboratory for 24 hours per ASTM C 192. The temperature at the center of one of the cylindrical specimens was recorded to establish the exact 24 hour curing history of the cylinder specimens themselves per ASTM C 1074 after which the temperature of the cylinders was assumed to be equivalent to the moist curing room at 23 C. Figure 8-3. Tim e/temperature history of 4x8 c ylinder s pecimens The early age temperature data shown in Figure 8 3 and 8 4 provide the recorded temperature history obtained from the 4x8 cylinder specimens and Sure Cure specimens, respectively. T he temperature history for each mixture was used to calculate maturity and equivalent age.

PAGE 119

119 Figure 8-4 Time/temperature h istory b lock Sure Cure c ylinder s pecimens Maturity Calculation The maturity of the concrete specimens for each mixture was calcu lated using the temperature history provided in Figures 8 3 and 8 4 incorporated into the Nurse Saul equation as outlined in ASTM C 1074, which states: ttTTM0 0)( (81) Where: M = maturity at age t T = average temperature of the concrete du ring the time interval t To = Datum temperature Equation 8 1 results in an increase of maturity only when the temperature of the concrete is higher than that of the datum temperature. The standardized test method for the calculation of the datum temperature is provided by ASTM C 1074 A1.1 and is discussed in detail in Chapter 4.

PAGE 120

120 Equivalent age calculation The equivalent age for the concrete specimens for each mixture was calculated using the temperature history, as provided in Figures 8 3 and 8 4 incorporated into the Arr henius equation as outlined in ASTM C 1074, which states: t ett TTR E er 0 273 1 273 1 (82) Where: te = equivalent age T = average temperature of the concrete during the time interval t Tr = reference temperature (typically 23C) E = activation energy, J/mol R = universal gas constant, 8.3144 J/(mol K) Revisiting Table 4 6, three Ea values were obtained for each mixture. The calculated Ea values are summarized in Table 8 2 and w ere used to calculate equivalent age for strength properties for normal cure specimens as compared to specimens match -cured with the Sure Cure System. Maturity calculations using ASTM 1074 (Nurse -Saul) are reported for normal versus Sure Cure for test only as represented by the first column of Ea values in table 8 2 (Ferraro and Tia, 2009). Table 8 -2. Summary of a pparent a ctivation e nergy per t esting and a nalysis m ethod Compressive strength of mortar cubes Isothermal Testing Mix Name Hyperbolic calculat ion (ASTM 1074) Exponential calculation Exponential calculation Mix 1 35642 37401 34235 Mix 2 33688 39932 50400 Mix 3 25757 20643 32982 Mix 4 30013 21158 37330

PAGE 121

121 Physical Data and Results Compressive Strength Testing Compressive strength vs. time Com pressive strength data was obtained at ages of 1, 2, 3, 7, 14 and 28 days to establish a rate of strength development over time which is often used to make determinations about the rate of hydration of the cementitious materials. It is often used as a meas ure of the quality of the concrete. Figure 8 5 is a graphical representation of strength vs. time for companion concrete specimens cast for each mixture. The results are consistent with typical findings where concrete with Portland cement alone gains stren gth fastest and where concrete with fly ash gains strength the slowest. Slag blend and ternary blend mixes gain strength at a rate between ordinary Portland cement concrete and concrete with fly ash, indicating that the partial replacement of Portland ceme nt with fly ash will reduce the initial rate of reaction of the concrete more than slag assuming the quantities have been replaced by the same mass. Specimen testing within this research was performed in accordance with ASTM C39. Figure 8-5. Compressive strength vs. time for each m ix

PAGE 122

122 Concrete specimens subjected to higher temperatures will gain strength at a faster rate than specimens that are exposed to lower temperatures (Carino and Tank, 1992; Carino, 2004; Mindess et. al, 2003). The strength vs. time results for concrete cylinder specimens composed of 100% Portland cement subject to standard curing (Figure 8 3) and the cylinders subjected to the Sure Cure / elevated temperature history (Figure 8 4) are shown in Figure 86. Figure 8-6. Compressiv e strength vs. time for Mix 1 100% Portland cement Compressive strength vs. maturity The strength maturity relationship was developed for the concrete specimens for each Mixture. The strength vs. time data for concrete cylinder specimens composed from Mix 1 subjected to standard curing and the cylinders subjected to the Sure Cure is provided in Figure 8 7. While a majority of the research comparing maturity and strength concludes that, for a given maturity, the strength will be the same regardless of t emperature history, the data indicates that the maturity -compressive strength relationship is not suitable for the concrete specimens composed of Portland cement.

PAGE 123

123 Possible reasons for the difference in strength vs. maturity can be attributed to the fact t hat research studying strength maturity relationships typically involves the isothermal curing of the concrete or mortar specimens prior to compressive strength testing. Another reason for the inconsistencies with the strength-maturity relationship developed for the concrete blocks was that testing was performed using mortar samples and there could be a slight difference between the maturity results for mortar and concrete. However, the most likely reason for the difference in the maturity results as show n in Figure 8 7 is the relatively unrefined nature of the Nurse Saul equations. Figure 8-7. Compressive strength vs. maturity for Mix 1 100% Portland cement Carino outlines several constraints using the Nurse Saul equation for the calculation of concr ete strength, which can be attributed to the possible deviation in the maturity-strength relationships. However, each of the constraints are properly addressed in the temperature histories for both sets of specimen curing history (Carino, 2004). In additi on, Carino later states that the maturity -strength relationship should only be considered an approximation, and that

PAGE 124

124 concrete does not have a unique strength -maturity relationship. The application of the maturity method for estimating in place strength is often abused or oversimplified by the use of the Nurse Saul relationship alone. As shown in Figure 8 7, the Nurse Saul relationship is not accurate for the Sure Cure vs. normal cure specimens for Mix 1. Appendix B provides similar inaccurate results for ea ch of the four block mixes using the Nurse Saul equation for the comparison of normal cure vs. Sure Cure maturity-strength relationships. Compressive strength vs. equivalent age A more accurate way to describe the strength of concrete with the incorporat ion of temperature history utilizes the concept of equivalent age. Using the Ea results provided by table 82 and the Arrhenius equation (8 2), the compressive strength versus equivalent age for the normally cured concrete specimens as well as the Sure Cur e specimens was calculated and is shown in Figure 8 8 (Ferraro and Tia, 2009). Figure 8-8. Compressive strength vs. equivalent age for Mix 1 100% Portland cement

PAGE 125

125 Three Ea values obtained for Mix 1 using the different testing and calculation technique s were used to create compressive strength versus equivalent age for the normal cure vs. Sure Cure specimen groups (cured at significantly higher temperatures as experienced by the large block specimen). The calculated results for compressive strength vs. equivalent age indicate there is not a practical variation with respect to the method used to obtain Ea. The compressive strength for the normal cure and the Sure Cure specimen groups are essentially the same for the first 8 days of curing history. Howeve r, the Sure Cure specimens exhibit higher compressive strength values after the 8 days. There are several factors contributing to this deviation in strength. The normal cured specimens were placed in a moisture room, which supplied additional curing water and also prevented water loss. However, the Sure Cure specimens were cured in a sealed condition to prevent moisture loss. Figure 89 is a photograph of a typical sure cure specimen mold. Figure 8 -9. Sure Cure specimen mold The Sure Cure specimen mold s are designed to provide insulation of the concrete as well as provide additional heat to the match cure specimen as programmed by the computer system. For this experiment, Sure Cure system was programmed to match the temperature produced by the portion of the section of the block as discussed in Figure 8 2. Since the Sure Cure molds are electrically powered, it is not feasible to add additional water to the Sure Cure specimens. Sealed curing of specimens can lead to self -desiccation of the concrete as hyd ration progresses (Mindess 2003, Neville 1995).

PAGE 126

126 Research has shown that specimens moist cured for 7 days then exposed to air will result in an immediate strength gain when compared to specimens continuously moist cured (Price 1951). Assuming all other con ditions are equal, the compressive strength of concrete cylinders of dry concrete will be higher than saturated concrete (Kesler, 1966). Decreased strength in moist as opposed to dry concrete has been attributed to the presence of moisture forcing the gel particles apart (Ross et. al 1996). Therefore, it can be assumed that the higher 8 day compressive strength associated with Sure Cure specimens is a result of the curing / saturation conditions rather than a temperature/hydration effect. The majority of the research regarding the maturity and equivalent age relationships focuses primarily on the study of ordinary concrete and mortar which utilize ordinary Portland cement as binder for compressive strength for specimens (Barnett et al., 2006). The study of the replacement of Portland cement with supplementary cementitious materials such as ground granulated blast furnace slag and fly ash has not been extensively researched. The equivalent age concept was applied to Mix 2, which replaced 50% of the Portlan d cement with granulated blast furnace slag. Using the data provided by table 8 2 and used in the Arrhenius equation (8 2), the compressive versus equivalent age for the normally cured concrete specimens as well as the Sure Cure specimens was calculated an d is shown in Figure 8 10. Unlike the compressive strength results obtained for Mixture 1, the normal cure and the Sure Cure specimen groups differ significantly for the entire curing history of the specimens. The observed difference is observed regardless of the testing method and calculation technique used to obtain Ea. The difference is due to the maturity or equivalent age maturity functions increase disproportionally at elevated temperatures and the increase depends upon the amount of supplementary cementitious material, the water to -cementitious material ratio, and the difference

PAGE 127

127 in temperature between the normal cured specimen and the temperature history of the large scale concrete. As a result, maturity relationships typically underestimate the strength development at elevated temperatures (Gajda, 2007). The data obtained for each of the mixes which incorporate supplementary cementitious materials clearly demonstrates this phenomenon. Figure 810. Compressive strength vs. equivalent age for Mix 2 50% Portland cement 50% s lag The value of Ea calculated by the isothermal calorimetry testing was the highest for each of the mixes with the exception of the Portland Cement only Mix. As a result, the strengthequivalent age relationships between the r egular cure and Sure Cure specimens were the most accurate for the compressive strength for the samples with large replacement of Portland cement with SCM. Figures 8 11 and 8 12 provide data showing the compressive strength versus the equivalent age for Mi x 3 (incorporating 35% replacement of Portland cement) and Mixture 4 which utilizes 30% slag and 20% fly ash as a replacement of Portland cement, respectively. The

PAGE 128

128 slag blend and ternary blend Mixes gain strength at a rate between ordinary Portland cement concrete and concrete with a high percentage replacement of flyash. Figure 811. Compressive strength vs. equivalent age for Mix 3 65% Portland cement 35% f ly ash Figure 812. Compressive strength vs. equivalent age for Mix 4 50% Portland ceme nt 30% slag 20% f ly ash

PAGE 129

129 Splitting Tensile Strength Testing Splitting tensile strength vs. time Tensile strength is particularly important for the potential model inputs since mass concrete failure typically takes place in tension. Tensile strength d ata was obtained at ages of 1, 2, 3, 7, 14 and 28 days to establish a rate of strength development over time, and Figure 8 13 is a graphical representation of the tensile strength vs. time for each mixture. The results are consistent with typical findings and compressive strength testing results, where concrete with Portland cement alone gains strength fastest and concrete with flyash gains strength the slowest. Slag blend, ternary Slag blend, and ternary blend mixes gain strength at a rate between ordinar y Portland cement concrete and concrete with a high percentage replacement of fly ash. The tensile strength testing of the cylinder samples within this research was performed in accordance with ASTM C496. Figure 813. Splitting tensile strength vs. time for each m ix

PAGE 130

130 Tensile strength vs. maturity The tensile strength / maturity relationship was developed for the concrete specimens for each Mixture. The strength vs. time data for concrete cylinder specimens composed from Mix 1 subjected to standard curing and the cylinders subjected to the Sure Cure is provided in Figure 8 14. While similar to the results obtained for the compressive strength testing, the maturity splitting tensile strength relationship is not, however, suitable for the concrete specimens c omposed of Portland cement. Appendix B provides similar inaccurate results for each of the four block mixes using the Nurse -Saul equation for the comparison of normal cure vs. Sure Cure maturity -strength relationships. Figure 814. Splitting tensile strength vs. maturity for Mix 1 100% Portland cement Tensile strength vs. equivalent age A relationship between splitting tensile strength/equivalent age was created using the same method used to create the compressive strength/equivalent age relationship. The splitting tensile strength vs. equivalent age results for the normally cured concrete specimens as well as the Sure Cure specimens was calculated and is shown in Figure 8 15 (Ferraro and Tia, 2009).

PAGE 131

131 The splitting tensile strength for the normal cure and the Sure Cure specimen groups is essentially the same for the entire curing history. The splitting tensile strength does not deviate between normal cure and Sure Cure after 8 days as observed during the compressive strength testing. This is because the fracture mechanisms of splitting tensile test specimens are generally the same regardless of moisture content. Research has shown that semi saturated and fully saturated concrete has lower strength than unsaturated concrete (Ross et. al. 1996). Since the normal cure concrete and Sure Cure concrete specimen groups were both at least semi -saturated, it is probable that the fracture mechanics for both specimen groups were identical. Figure 815. Splitting tensile strength vs. equivalent age for Mix 1 100% Portland cement Since it is known that mass concrete will typically crack under tensile forces due to thermal gradients, it is particularly useful to consider splitting tensile testing for the model input parameters. The relationship between splitting tensile testing strength versus equivalent age for concrete composed of Portland cement can be accurately predicted regardless of temperature history.

PAGE 132

132 The equivalent age concept was applied to Mix 2, which replaced 50% of the Portland cement with granulat ed blast furnace slag. The strength versus equivalent age results for the normally cured concrete specimens as well as the Sure Cure specimens was calculated and is shown in Figure 8 16. Figure 816. Splitting tensile strength vs. equivalent age for Mix 2 50% Portland cement 50% s lag Unlike the compressive strength results obtained for Mixture 1, the normal cure and the Sure Cure specimen groups differ significantly for the entire curing history of the specimens. This observed result is consistent with the data obtained from compressive strength testing with respect to equivalent age. Again, the equivalent age maturity functions increase disproportionally at elevated temperatures (Gajda 2007). The data obtained for each of the Mixes which incorporate s upplementary cementitious materials clearly demonstrate this phenomenon. Figures 816 and 8 17 show the splitting tensile strength versus the equivalent age for Mix 3 (35% replacement of Portland cement) and Mixture 4 which utilizes 30% slag and 20% fly a sh as replacement of Portland cement, respectively.

PAGE 133

133 Figure 817. Splitting tensile strength vs. equivalent age for Mix 3 65% Portland cement 35% f ly ash Figure 818. Splitting tensile strength vs. equivalent age for Mix 4 50% Portland cement 30% slag 20% f ly ash The splitting tensile strength testing vs. equivalent age results for Mix 3 show relatively similar values between the normal cure and Sure Cure specimens for the first 150 hours of

PAGE 134

134 testing. However, at later ages the results deviate and the underestimation of strength with the normal cure specimens occurs. The value of Ea calculated by the isothermal calorimetry test was the highest for each of the Mixes with the exception of the Portland Cement only Mix. Similar to the compressive s trength testing, the tensile strength -equivalent age relationships between the regular cure and Sure Cure specimens were the most accurate per the Ea obtained by isothermal calorimetry for the Mixes containing large replacements of SCM. Revisiting Figure 8 11, the results for the compressive strength and equivalent age for Portland cement with fly ash are somewhat closer in strength than the mix which incorporates blast furnace slag or the ternary blend (Figures 8 10 and 8 12). A possible reason for the relatively consistent result is that the percentage of Portland cement replacement is lowest in Mixture 3 (35%) However, it is mostly likely that the replacement of Portland cement with slag results in the disproportional increase in strength and physical properties at elevated temperatures. Compressive Modulus of Elasticity Most structures are designed to undergo relatively small in-service deformations. Typically, the primary failure mechanism of early age mass concrete is a result of thermal strain exceed i ng the strain capacity. T he cracking of the concrete occurs as a result Therefore, it is necessary to quantify the linear elastic properties of the concrete while under compression and tension. The linear relationship between the stress and the strain of a material is known as Hookes Law (Beer and Johnson 1992) and is expressed: E (83) Where: = the stress of the material E = is Young's Modulus = linear strain

PAGE 135

135 However, concrete behaves as a nonlinear inelastic material in both compression and in tension, therefore a single linear value modulus for elasticity does not accurately describe the stress/strain behavior of concrete (Mindess et. al 2003). Figure 8 19 is an illustration of the various points within the stress strain curve. Figure 819. Typical stress -strain diagram for concrete, showing the different elastic moduli (Mindess 2 003) Tangent modulus Tangent modulus is defined as the slope of the stress/strain curve at any given point. It can also be considered the instantaneous rate of change of the stress with respect to strain (Philleo, 1966). The closest approximation to the modulus of elasticity as described by Hookes law can be described as the initial tangent modulus (also referred to as dynamic modulus), which is the slope of the tangent to the stress -strain curve at the origin (Mindess et al 2003, Mehta 1986). Both tangent and initial tangent moduli are depicted in Figure 8 19.

PAGE 136

136 Secant modulus The secant modulus is the slope of the secant between the origin and a given point on the stress/strain curve. It is dependent upon the intersecting point on the stress/strain curve. The secant modulus is considered to be the instantaneous rate of change of the stress with respect to strain (Philleo, 1966). Typically, the secant modulus is determined to be the secant between the origin and the point on the stress/strain curve which corresponds to 40% of the failure stress (Mehta, 1986). Chord modulus The chord modulus is the slope of a line drawn between any two points on the stress/strain curve. Research has shown that the initial tangent and secant moduli can produce errors in m easurement due to either initial loading seating effects or defects in the specimen both of which result in nonlinearity at the early portions of the stress / strain curve (Mindess et al, 2003). As such, the Standard Test Method for Static Modulus of Elasti city of Concrete in Compression (ASTM C 469) specifies the chord modulus be drawn between points corresponding to 50 microstrain and 40% of the ultimate strength. Modulus of elasticity methodology As concrete hydrates, the modulus of elasticity increases, and is analogous to the other strength properties of concrete. However, the increase in elastic modulus of elasticity is detrimental to early age mass concrete as the propensity for thermal cracking increases with increased modulus. Therefore, the most co nservative estimate of modulus of elasticity for mass concrete is the highest calculated value per the stress/strain relationship. The modulus of cylinder samples tested within this research was performed in accordance with ASTM C469. Prior to the modulus of elasticity testing, the ultimate load was obtained using the average compressive strength per ASTM C39. ASTM C 469 states that the specimens are to

PAGE 137

137 be tested at stresses of up to 40% of ultimate load. Figure 8 20 is a stress -strain curve of a typical c oncrete 4x8 cylindrical specimen tested in accordance with ASTM C 469. Figure 8 20. Stress -strain diagram for concrete specimen Four different calculations for elastic modulus are presented in Figure 8 20: the initial tangent modulus, the secant modulus, the chord modulus per ASTM C 496, and a truncated secant modulus. 1. The initial tangent modulus was calculated as the best fit line between the first several points of data. The modulus was obtained from the line drawn through the origin and a point on the stress -strain curve at 2 microstrain. The initial tangent modulus is 4,551,135 psi for this specimen. 2. The secant modulus as prescribed by Mehta (1986) is a line between the origin and the corresponding strain at 40% of the ultimate load. The secant modulus is calculated to be 1,933,840 psi using this technique. 3. The chord modulus per ASTM C 469 was calculated to be 1,897,142 psi.

PAGE 138

138 Research has shown that microcracking can begin in concrete specimens when subjected to loads of between 30% and 40% of ultimate load (Zheng, 2007, Bascoul 1996, Hsu et al, 1963). The motivation behind the utilization of 40% of ultimate load for the standardized calculation of chord modulus is to obtain a modulus of elasticity of concrete materials at design loads (Ke sler 1966). The standard method for creep of concrete in compression (ASTM C 512, 2008) specifies 40% of the ultimate load for testing. Per this test method, it is expected that the concrete will experience permanent deformations as a result of being subjected to such loads for long periods of time. Therefore, it is appropriate to conclude that concrete will experience micro-cracking or micro -damage at loads that are 40% of ultimate. The onset of micro -cracking within the test specimen will result in perm anent deformation and lower modulus which for the purposes of this researchis not conservative. Therefore, a new method for the calculation of modulus of elasticity is being proposed by this research. The truncated secant modulus is a line drawn betw een the origin and the corresponding strain at 30% of the ultimate load. The truncated secant modulus shown in Figure 8 20 is 1,980,890 psi, which is the largest of the static moduli. The compressive modulus of elasticity was calculated using the truncated secant modulus for each of the specimens used in this research. Compressive modulus of elasticity vs. time Modulus of elasticity data was obtained at ages of 1, 2, 3, 7, 14 and 28 days to establish a rate of modulus development over time for each of the concrete Mixtures. Figure 8 21 is a graphical representation of the modulus of elasticity vs. time for companion concrete specimens cast for each mixture. The results are consistent with typical findings where concrete with Portland cement alone gains str ength fastest and concrete with flyash gains strength the slowest.

PAGE 139

139 Concrete mixes which replace Portland cement with slag, fly ash, or both (ternary blend Mixes) gain strength at a rate between ordinary Portland cement concrete. Figure 8 21. Compressi ve modulus o f elasticity vs. time for each mix Compressive modulus of elasticity vs. maturity The modulus of elasticity / maturity relationship was determined for each mixture. The modulus of elasticity vs. time result for Mix 1 subjected to standard curi ng and the cylinders subjected to the Sure Cure is provided by Figure 8 22. These results are similar to those obtained for the compressive strength and tensile strength testing. Thus, the maturity-modulus of elasticity relationship is not suitable for the concrete specimens composed of Portland cement. Appendix B provides similar inaccurate results for each of the four block Mixes using the Nurse Saul equation for the comparison of normal cure vs. Sure Cure maturity modulus relationships.

PAGE 140

140 Figure 8 22. Compressive modulus of elasticity vs. maturity for Mix 1 100% Portland cement Compressive modulus of elasticity vs. equivalent age A relationship between modulus of elasticity and equivalent age was developed using the same method as the modulus of el asticity equivalent age relationship. The modulus of elasticity versus equivalent age for the normally cured concrete specimens as well as the Sure Cure specimens was calculated and is shown in Figure 823. The modulus of elasticity for the normal cure an d the Sure Cure specimen groups are essentially the same for the entire curing history. The modulus of elasticity does not deviate between normal cure and Sure Cure after 8 days as observed with the compressive strength testing. This can be attributed to t he compressive loading for the modulus of elasticity testing remaining below the threshold for micro cracking. By design, the loading remained in the elastic range of the concrete. Therefore the fracture mechanics that apply to compressive strength testing do not apply to modulus of elasticity testing.

PAGE 141

141 Figure 823. Compressive modulus of elasticity vs. equivalent age for Mix 1100% Portland cement Since it is known that mass concrete will typically crack under tensile forces due to thermal strains, it is particularly useful to consider modulus of elasticity testing for the model input parameters. Modulus of elasticity versus equivalent age for concrete composed of Portland cement as a relationship is accurately predicted regardless of temperature hist ory. The equivalent age concept was applied to Mix 2 which replaced 50% of the Portland cement with granulated blast furnace slag. The equivalent age vs. strength result for the normally cured concrete specimens as well as the Sure Cure specimens was calculated and is shown in Figure 8 24. Unlike the compressive strength results obtained for Mixture 1, the normal cure and the Sure Cure specimen groups differ significantly for the entire curing history of the specimens. This observed result is consistent wi th the data obtained from compressive strength and tensile strength testing with respect to equivalent age. Again, the equivalent age maturity functions increase disproportionally at elevated temperatures (Gajda, 2007). The data obtained for each of

PAGE 142

142 the mixes which incorporate supplementary cementitious materials clearly demonstrate this phenomenon. Figures 8 25 and 8 26 show the modulus of elasticity versus the equivalent age for Mix 3 incorporating 35% replacement of Portland cement and Mixture 4 utiliz ing 30% slag and 20% fly ash as replacement of Portland cement, respectively. As with the results obtained from the splitting tensile strength testing, the modulus of elasticity vs. equivalent age for Mix 3 shows relatively similar values between the norma l cure and Sure Cure specimens for the first 150 hours of testing. However, at later ages the results deviate and the underestimation of strength with the normal cure specimens occurs. Revisiting Figure 8 11, the equivalent age model for the compressive st rength versus equivalent for the mixes which do not contain large amount of slag replacement, more accurately model the evolution of physical and strength properties than those mixes which incorporate blast furnace slag. A possible reason for the relativel y consistent data is that the percentage of Portland cement replacement is lowest in Mixture 3 (35%). The relationships developed with respect to modulus of elasticity equivalent age shown in Figures 8 24 and 8 25 are consistent with the relationships deve loped for tensile strength testing. Figure 8 24. Compressive modulus of elasticity vs. equivalent age for Mix 2 50% Portland cement 50% s lag

PAGE 143

143 The value of Ea calculated by the isothermal calorimetry test was the highest for each of the mixes with the exception of the Portland Cement only Mix. Similar to the results obtained by the compressive and tensile strength testing, the modulus -equivalent age relationships between the regular cure and Sure Cure specimens were the most accurate. Figure 8 25. Compressive modulus of elasticity vs. equivalent age equivalent age for Mix 3 65% Portland cement 35% f ly ash Figure 8-26. Compressive modulus of elasticity vs. equivalent age for Mix 4 50% Portland cement 30% slag 20% f ly ash

PAGE 144

144 Tensile Modulus of Elasticity The primary failure mechanism of mass concrete can be attributed to the tensile stresses induced by thermal gradients. Therefore, the tensile modulus is of the utmost importance when considering the physical properties of concrete. There is very little research available which compares tensile modulus to compressive modulus. The majority of tensile testing performed on concrete materials is performed via indirect methods such as the splitting tensile test discussed earlier in this chapter. The re is no standardized method for the direct tensile testing of concrete for strength or modulus. Therefore, it was decided to use the existing standardized flexural test per ASTM C 78 to measure modulus of elasticity. The test was modified to incorporate l inear voltage displacement transducers (LVDTs) as shown in Figure 8 26. Equation 83 is a calculation of modulus incorporating the measured load and deflection. I PL MOE 28 2/ (83) Where: MOE = Modulus of Elasticity (psi) P= Load (lb) = Average deflection (in) I = Moment of inertia = bh3 (in4) Figure 827. Photo of tensile (f lexural) m odulus of e lasticity test configuration

PAGE 145

145 It may be more appropriate for the tensile testing presented herein to be considered flexural modulus of elas ticity testing. However, the stress -strain diagram produced for a concrete beam subjected to a flexural load is due to tensile stresses. Furthermore, the third -point bending test imparts a uniform moment across the loading area of the test specimen which f acilitates calculation of the modulus of elasticity. Tensile modulus of elasticity vs. time Compressive strength data was obtained at ages of 1, 2, 3, 7, 14 and 28 days to establish a rate of strength development over time, which is considered to be ana logous to the rate of hydration of the cementitious materials over time. Figure 8 28 is a graphical representation of the tensile modulus of elasticity vs. time for each mixture. The results are consistent with findings per the compressive strength, splitt ing tensile strength and modulus of elasticity testing results, where concrete with Portland cement alone gains strength fastest and concrete with flyash gains strength the slowest. The slag blend and ternary blend Mixes gain strength at a rate between or dinary Portland cement concrete and concrete with a high percentage replacement of flyash. Figure 8 28. Tensile modulus of elasticity vs. time for each mix

PAGE 146

146 Comparison of t ensile and compressive m odulus of e lasticity The moduli of elasticity / maturity r elationships were not developed for the concrete specimens for each mixture. The Sure Cure system is not capable of match curing the typical prism specimens (6x6x22) required for flexural/modulus testing. Appendix B provides a data summary of the tensil e modulus of elasticity results for each of the four block Mixes using the Nurse Saul equation. Equivalent age relationships were developed for the laboratory cured specimens for both the prism specimens and the cylinder specimens. The temperature history data for the 4x8 cylindrical specimens presented in Figure 8 3 was used to establish the equivalent age of the specimens and was recorded to establish the exact curing history and equivalent age of the cylinder specimens. Accordingly, the same procedure was performed on the prism specimens used to establish the equivalent age. There is little research available regarding the comparison of tensile and compressive modulus of elasticity data, primarily because the determination of the modulus of elasticity in tension is rarely performed. As previously discussed, tensile testing of concrete materials is performed indirectly. Therefore, the modulus data obtained via an indirect test is difficult to obtain and there is currently no standardized method available to acquire tensile modulus of elasticity of concrete. Balendran (1995) performed modulus of elasticity testing on concrete materials in direct tension using a beam specimen in which the central portion had a reduced cross -sectional area. In his experiment, Ballendran compared the tensile vs. compressive modulus of several cementitious mixtures including concrete containing limestone coarse aggregate. However, the experiment was limited in that the concrete was tested at an age of 28 days only. His results indicated that the modulus of elasticity in compression is slightly larger (3%) than the modulus

PAGE 147

147 of elasticity in tension. He concluded that the modulus in tension is nearly equal to the modulus in tension at a stress level of one third the ultimate stre ss (Balendran, 1995p 16). The tensile and compressive modulus of elasticity vs. equivalent age for the four mixes are presented in Figures 8 29 through 8 32. The results indicate that the compressive modulus of elasticity is approximately equal to th e tensile modulus for each of the mixes. However, data obtained for the compressive modulus of elasticity is slightly lower than the tensile modulus at earlier ages (one day). Conversely, the compressive modulus of elasticity is significantly larger (15%) than the tensile modulus of elasticity at 28 days. This difference may be attributed to the difference in testing equipment (stress and strain measurements). The stress strain curves were obtained in the same manner as Balendran, (using loads less than 3 3% of ultimate), and thus the similarity in the results indicates the tensile and compressive stresses can be considered the same value for prediction and modeling purposes. Figure 8 29. Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 1

PAGE 148

148 Figure 8 30. Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 2 Figure 8 31. Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 3

PAGE 149

149 Figure 8 32. Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 4 Summary of Findings The physical testing of properties the concrete in this study resulted in the following: 1. The application of Nurse -Saul mat urity method is to determine physical and strength parameters based on the time temperature history of concrete. The Nurse Saul per ASTM 1074 does not accurately model the physical and strength properties of concrete with different time -temperature histories of concrete. 2. The equivalent age method per ASTM C1074 accurately models the evolution of physical and strength parameters with different time temperature histories time temperature history which can be used for the estimation of concrete created with Portland cement. 3. Neither the Nurse -Saul equation nor the equivalent age method per ASTM C1074 accurately models the hydration and strength evolution of Portland cement concretes with large portions of supplementary cementitious materials with different cu ring histories. Further research is needed to create maturity/equivalent age models for concrete using large proportions of fillers (Ferraro and Tia, 2009). 4. The modulus of elasticity of concrete material is most accurately measured -with comparable result sbetween different physical tests (compression, tension) at stress levels equal to or less than 30% of ultimate.

PAGE 150

150 CHAPTER 9 LARGE -SCALE BLOCK EXPERIME NT THERMAL TESTING The testing of the early age concrete subsequent to the creation of the large sca le block specimens was performed using thermal laboratory testing methods. Results obtained from laboratory testing of the concrete obtained from the same mixes as the large -scale block specimens will serve as potential parameters the prediction of the behavior of mass concrete. Coefficient of Thermal Expansion Testing The coefficient of thermal expansion (CTE) is a very important thermal property with regard to the prediction of the behavior of early age mass concrete. Virtually all materials experience vo lumetric changes when subjected to temperature change. Concrete behaves like most other elastic materials in that it expands volumetrically when it is subjected to temperature increase and shrinks volumetrically when it is subjected to temperature decrease. The magnitude of the volumetric change is typically quantified by the CTE. The standardized test method for CTE is defined as the linear change in material with respect to a change in temperature: T L L CTEO a (9-1) Where: La = actual length change of specimen during temperature change LO = measured length of specimen at room temperature = measured temperature change The coefficient of thermal expansion was obtained each of the four mixes at ages of 1, 2, 3, 7, 14 and 28 days. The testing was performed in accordance with AASHTO TP 60 with the following exceptions: 1. 7.1 The early age specimens (1, and 2 day) were not conditioned by submersion in lime water for 48 hours prior to testing.

PAGE 151

151 2. 8.3 Only two test segments for each test were performed at early ages because each test segment required 8 hours (16 hours total). One segment was the 50C temperature cycle; the other segment was the 10C temperature cycle 3. Individual specimens were tested once in an effort to avoid mod ifying the equivalent age of specimens at each age Figure 9 1 is a graphical representation of the coefficient of thermal expansion vs. time for each mixture. The results indicate that the values of coefficient of thermal expansion obtained at 1,2, and 3 days of age were slightly lower than later age values for each mix. However, after seven days, the values of coefficient of thermal expansion slightly increased for the blocks containing supplementary cementitious materials. The increase of CTE values of the concrete at early age per obtained in this study is similar to the findings by Cusson and Hoogeveen (2006), where the CTE at early ages is less than at later ages. Figure 9 -1. Coefficient of thermal expansion vs. time for each mix Companion specime ns for CTE testing using the Sure Cure system were not created due to the limited number of 4x8 specimen molds. CTE testing was performed primarily to obtain

PAGE 152

152 values for use in potential analytical models and to record early age values of CTE vs. time f or concrete made with materials from Florida. Currently, the standardized method for the measurement of CTE (AASHTO TP 602007) has been withdrawn due to the potential concerns with the calibration factor (of stainless steel) used to calibrate the test (C rawford, 2009). Therefore, the results obtained under this research were calculated per the calibration testing performed at the FDOT SMO. Thermal Diffusivity Testing The quantification of the thermal diffusivity of concrete is necessary for the prediction of the behavior of mass concrete. The thermal diffusivity is measure of the rate at which the temperature changes within concrete (Mitchell, 1966). Knowledge of the rate at which concrete is able to absorb or dissipate heat is important in determining th e rate of cooling of mass concrete, assuming an elevated early age temperature has been obtained. The diffusivity is of particular importance when calculating the temperature differential of concrete at early ages. The equation used to calculate thermal di ffusivity is determined from equation 9 2. pC k a Where: a= thermal diffusivity (m2/day) k = thermal conductivity (kJ/mCh) Cp = specific heat capacity (kJ/kgC) = concrete density (kg/m3) The standardized test for the therma l diffusivity of concrete is outlined in CRD 36 (1973). The test involves the temperature monitoring of a 6x12 cylindrical concrete specimen as it is placed from one high temperature water bath to another lower temperature water bath. The testing method requires that the concrete be immersed in a water bath at 100C until the center of

PAGE 153

153 the specimen reaches temperature equilibrium; then the specimen is immersed in a low temperature water bath while the time temperature history of the specimen is recorded. The procedure for this research was slightly altered to accommodate the necessary parameters per the testing of concrete specimens at early ages. In an effort to avoid heating (and thereby artificially increasing the maturity of the concrete specimen) at early ages, the specimens were left in the standard moist cure room at 23C and immersed in a hot water bath for testing, which is essentially the reverse of what the standard test method specifies. Ochiai (2009) performed testing to determine the potentia l difference between concrete specimens initially hot (CRD 36) and initially cold (modified CRD 36) and found the difference between the obtained values was found to be negligible. Figure 9 -2. Thermal diffusivity vs. time for each mix The early age ther mal diffusivity results obtained by Ochiai for the concrete specimens for each mix are presented in Figure 9 2. The results indicate that the thermal diffusivity increases with respect to time for each of the concrete mixes. The concrete mixes that contain large

PAGE 154

154 replacements of Portland cement with blast furnace slag or fly ash have a lower thermal diffusivity. Accordingly, the concrete which contains large replacements of amounts of fly ash and slag should be considered to have lower diffusivity values at early ages. ACI 207.2 lists the thermal diffusivity of concrete in which the values of thermal diffusivity are reported. The diffusivity values for concrete made with limestone coarse aggregate, reported by Table 3.1 of ACI 207.2 range from 0.029 0.054 (m2/hour). The results obtained from this study, are significantly lower than the values provided ACI 207.2. A potential reason for the difference in diffusivity values between the concrete tested in this study and the values published in ACI 207.2 can be attributed to the difference in coarse aggregate used in the concrete. The coarse aggregate used in this study was limestone produced from Florida, which is more porous, less dense, and softer than limestone produced from other areas of the U.S. To differ entiate between limestone produced in Florida, and limestone produced elsewhere, Florida limestone has been called Florida limerock by the local concrete industry. The porosity and resultant density of the lime rock used for coarse aggregate used in this study is the cause for the lower reported diffusivity values as compared to those reported by ACI 207. Tyner (1946) reported lower thermal conductivity (which is similar to diffusivity) with concrete made with Florida limerock as compared to concrete made with limestone from Minnesota. The data obtained in this study per Figure 9 2 indicates that the thermal diffusivity increases with concrete age. The results obtained in this study are similar to the results published by Hansen et al (1982) and Morabito (2001). The results are contrary those reported by DeSchutter and Taerwe (1995) who reported a decrease in thermal diffusivity of concrete with

PAGE 155

155 respect to age. However, no consensus has been reached regarding the increase of decrease of the thermal diffusi vity at early ages (DeSchutter and Taerwe, 1995, Chirdon et. al 2007). One potential explanation for the contrasting results between the diffusivity testing performed in this study and the results reported DeSchutter and Taerwe is most likely due to the different curing methods used in each respective experiment. The specimens used in this study were moist cured in accordance with ASTM C192. However, DeSchutter and Taerwe sealed their specimens which would allow for natural desiccation to take place. Therefore, the diffusivity results over time would decrease proportionally. As concrete hydrates and the microstructure becomes more dense, the absorptive, and diffusive properties decrease accordingly. Since the thermal diffusivity test measures the rate of temperature change of a concrete specimen submerged in water, a saturated sample (with available pore water in the structure) would be more likely to conduct and transfer heat from than an unsaturated specimen. Tyner (1946) found that the thermal diffusivity of concrete is dependent upon the moisture content of concrete made with Florida lime rock. Therefore, the contrasting results obtained by this study and DeSchutter in Taerwe is most likely due to specimen conditioning. A potential reason for the increa se in the thermal diffusivity of concrete per the results in this study could be due to the formation of the concrete microstructure which would include the presence of a saturated void system, as a result of the addition of curing water to the concrete. Specific Heat Capacity Testing The quantification of the thermal specific heat capacity of concrete is necessary for the prediction of the behavior of mass concrete. The specific heat capacity is the measure of energy or heat required to change unit mass by a unit temperature. As previously stated, the thermal properties of Florida limerock are different than the values for limestone. Thus, in an effort to

PAGE 156

156 obtain the specific heat of concrete which incorporates limerock, it was decided to measure the spec ific heat of the materials used in this study. Ochiai (2009) discusses in detail the testing procedure developed and performed for the measurement of the specific heat capacity of the cement paste, coarse aggregate, fine aggregate and concrete materials u sed in this study. Table 91 provides the results obtained for the average specific heat capacity of the materials used in this study. The specific heat capacity of the cement paste and concrete was measured at an age of 1 day. Table 9 -1. Specific h eat f or c oncrete m aterial Material Type Average Specific Heat Capacity (kJ/ kg C ) Cement Paste 1.10 Fine Aggregate 0.84 Coarse Aggregate 0.9 1 Concrete 1.19 The results indicate that the specific heat capacity testing method may have a potential inconsis tency because the results obtained for the concrete specimen are larger than any of the constituent materials. However, another explanation for the larger value of the specific heat capacity of concrete compared to its constituent materials could be attrib uted to the fact that concrete has entrapped air and a different capillary system when compared to cement paste. Therefore, the results obtained in this study are most likely accurate. ACI 207.2 lists the values of the specific heat capacity of concretes made with various coarse aggregates. The values of specific heat capacity of concrete made with limestone coarse aggregate, reported by Table 3.1 of ACI 207.2 range from 0.920 1.033 (kJ/ kg C ). T he value of specific heat capacity of concrete measured in this study, (1.19 kJ/ kg C) is significantly larger than the values provided ACI 207.2. The most likely reason for the difference in diffusivity values between the concrete tested in this study and the values published in ACI 207.2

PAGE 157

157 is due to the difference in coarse aggregate used in the concrete. Referring to equation 9 1, the thermal diffusivity is inversely proportional to the specific heat capacity. Therefore, the values for the thermal diffusivity testing and the specific heat capacity testing are c onsistent despite the fact that they were conducted independently. However, further research should be performed which involves the measurement of specific heat capacity of different concrete and constituent materials for the making of concrete. Summary of Findings The laboratory testing and of the thermal properties of concrete resulted in the following: 1. The coefficient of thermal expansion of concrete is lower at early ages than at later ages. 2. Concrete which contains large percentages of supplementary cementitious material replacement of Portland cement has lower coefficient of thermal expansion values than concrete made with Portland cement alone. 3. Concrete made with Florida limerock does not have the thermal properties as concrete made with limestone not local to Florida. 4. The thermal diffusivity of saturated concrete increases at early ages. 5. The thermal diffusivity of concrete made with Florida limerock is lower than concrete. made with limestone not local to Florida 6. The specific heat capacity of concrete made with Florida limerock is higher than concrete made with limestone not local to Florida.

PAGE 158

158 CHAPTER 10 RECOMMENDED TESTING PROGRAM FOR MASS CON CRETE The primary objective of this research was the development of a laboratory testing program to predict the behavior of mass concrete. Accordingly, this research was performed to determine which laboratory tests were most appropriate to be used as a predictor of the thermal and physical behavior of concrete. Furthermore, conducting this research establi shed some of the physical properties necessary for the prediction of mass concrete and which, therefore do not need to be evaluated for a given mixture design. Recommended Laboratory Testing Method for Measurement of Heat Generation The measurement of hea t generation and resultant temperature rise of concrete is perhaps the most important parameter for the prediction of the behavior of mass concrete. Temperature rise, relative strength properties, and resultant maturity/equivalent age functions are all dep endent upon the heat generated by concrete. Three calorimetry methods were investigated for the measurement of heat generation in concrete materials: isothermal conduction calorimetry, semi adiabatic calorimetry, and Sure Cure/adiabatic calorimetry. Of the three methods investigated, the isothermal calorimetry method was determined to be the most appropriate method for the quantification of heat generation of cementitious materials at early ages. There are several reasons as to why the isothermal calorim etry method was favored over the other calorimetry techniques. First, the isothermal calorimetry method provided the most accurate and repeatable results for the measurement of the heat generation of cementitious materials. Second, the results provided by the isothermal calorimetry are in units of power (mW/g) and energy (J/g). Although this research established some baseline values for the specific heat capacity of the ingredients used to make concrete, there are potential errors in those values which coul d provide erroneous results in the conversion from temperature rise as

PAGE 159

159 measured by semi adiabatic and adiabatic calorimetry. Third, the isothermal conduction calorimeter is commercially available and standardization of the test method for the determinatio n of energy rise of cementitious materials is imminent. Finally, there have been some recent developments in modeling/prediction software programs which utilize the energy rise data for cementitious materials for input parameters. Therefore, the use of the isothermal conduction calorimeter is recommended for the measurement of heat/energy rise of concrete for the purposes of prediction of the behavior of mass concrete. Recommended Laboratory Testing Method for Measuring Maturity/Equivalent Age The laboratory testing method for the establishment of maturity and equivalent age relationships based on the time temperature history falls under ASTM 1074. The relative strength relationships developed by the maturity method (ASTM 10746.2) also known as the Nurse Saul maturity method did not yield accurate results. Therefore, it is not recommended to use the Nurse -Saul method for the prediction of strength properties of mass concrete. The standardized equivalent age method developed relative strength relationship s per ASTM 10746.3 (also known as the Arrhenius equation) which utilized a hyperbolic mathematical model for the evolution of compressive strength of mortar cubes. This research investigated a second equivalent age method (Schindler, 2003) which utilized the compressive strength of mortar cubes as well, but developed an exponential mathematical model other than the model than the one prescribed in ASTM 1074.6.3 (hyperbolic method) for the establishment of Ea. A third method for the establishment of equiva lent age (and Ea) was investigated utilizing isothermal calorimetry of cementitious material in lieu of compressive strength testing of mortar cubes. As previously stated, the isothermal calorimetry test is required for the determination of the measurement of heat generation characteristics of the cementitious materials. Accordingly it is possible to use the data obtained from isothermal calorimetry testing (for heat generation) to

PAGE 160

160 determine the Ea of the cementtious material. Therefore, the compressive str ength testing per ASTM 1074 would not be needed for the determination of Ea, if the isothermal calorimetry testing was used in its place. Each of the three equivalent age techniques produced accurate results for the prediction of the relative strength of concrete which utilized Portland cement as the only cementitious material. However, initial testing indicated that the Ea results obtained by the isothermal testing method were the most accurate for cementitious materials in which Portland cement was repla ced with a large percentage of supplementary cementitious material (35 50%). However, the current models established for equivalent age relationships for concrete with large amounts of supplementary cementitious materials do not adequately predict the beha vior of relative strength of concrete. Therefore, the equivalent age relationships should be used to predict the strength behavior of concrete. Following the results obtained in this research, the Ea as calculated via isothermal calorimetry testing and th e standardized test method may be used for the prediction of concrete with large replacement of SCM. However, the models do not adequately predict the behavior of relative strength of concrete with large replacements of SCM. Recommended Laboratory Testing Methods for Strength and Modulus of Elasticity Compressive strength The most commonly measured property of hardened concrete is compressive strength. The FDOT specification requires a standardized minimum compressive strength, at an age of 28 days of all structural concrete used in the State of Florida. Thus, the compressive strength has become the fundamental strength test for concrete. Despite the fact that mass concrete will rarely experience failure in a compressive manner, it is recommended to perform compressive strength testing of concrete for acceptance and comparative purposes.

PAGE 161

161 Compressive M odulus of Elasticity The modulus of elasticity is the most important strength parameter for the prediction of the behavior of mass concrete. Near surface tensi le strains (as predicted by CTE measurements) due to the expansion of the hotter inner core of mass concrete can be used to calculate tensile stress if the modulus of elasticity is known. Therefore, the modulus of elasticity is an important parameter for t he prediction of potential thermal/tensile cracking of mass concrete. The compressive modulus of elasticity test is the only standardized test for measurement of the modulus of elasticity for concrete. The findings of this research indicate that the compr essive modulus of elasticity is approximately the same as the tensile (flexural) modulus of elasticity when tested at loads of 33% of ultimate. Therefore, the modulus of elasticity measured compressively can be considered the same value for the tensile mod ulus of elasticity for prediction and modeling purposes of mass concrete. Therefore, it is not necessary to determine tensile/flexural modulus of elasticity via laboratory testing. Tensile Strength Tensile strength is particularly important for the predi ction of mass concrete since the failure of mass concrete at early ages is typically tensile in nature. Near surface tensile stresses (as predicted by CTE) and modulus of elasticity measurements of mass concrete (as predicted by modeling software) can be compared with tensile strength obtained via laboratory testing. Accordingly, the prediction of cracking can be established with respect to the expansive nature (due to temperature rise or temperature differentials) within the mass concrete. When the predict ed tensile stress exceeds the ultimate tensile stress of the concrete obtained by laboratory testing cracking will occur.

PAGE 162

162 Recommended Laboratory Testing Method for Measurement of Coefficient of Thermal Expansion The coefficient of thermal expansion of conc rete is perhaps the most important parameter to accurately quantify with respect to mass concrete. Currently, the standardized method for the measurement of CTE (AASHTO TP 602007) has been withdrawn due to the potential concerns with the calibration facto r (of stainless steel) used to calibrate the test (Crawford, 2009). Therefore, the results obtained under this research were calculated per the calibration testing performed at the FDOT SMO. Considering the recent developments and potential problems regard ing the CTE test method, it is recommended to perform CTE testing in the laboratory for the prediction of mass concrete. Furthermore, it should be noted that earlyage CTE is a necessary parameter to properly establish and consider its variability in the p rediction of the behavior of mass concrete. Recommended Physical Parameter for the Concrete Diffusivity The thermal diffusivity testing reported in this research as performed by Ochiai (2009) indicates that the thermal diffusivity values for concrete at early ages range from 0.513ft2/day to 0.748ft2/day whereas the thermal diffusivity of concrete made with limestone aggregate as reported by ACI 207.2 is 1.22ft2/day. The reason for the difference in the diffusivity values obtained in this research and thos e obtained by ACI 207.2 is due to the coarse aggregate within the concrete. Therefore, the values for thermal diffusivity reported in this research should be used for the prediction of mass concrete which incorporate coarse aggregate from Florida. However, for concrete created with coarse aggregate from elsewhere, the values per ACI 207.2 should be considered for the prediction of the behavior of mass concrete.

PAGE 163

163 Recommended Physical Parameter for the Specific Heat Capacity The specific heat capacity testing reported in this research as performed by Ochiai (2009) indicates that the specific heat capacity values for concrete at early ages range from 1.1 to 1.2 J/(g -C), whereas the specific heat capacity of concrete made with limestone aggregate as reported by ACI 207.2 is 0.91 to 1.05 J/(g-C). The reason for the difference in the diffusivity values obtained in this research and those obtained by ACI 207.2 is due to the coarse aggregate within the concrete. Similar to the results obtained for thermal diffusivi ty, the specific heat capacity values reported in this research should be used for the prediction of mass concrete which incorporate coarse aggregate from Florida. However, for concrete created with coarse aggregate from elsewhere, the values per ACI 207.2 should be considered for the prediction of the behavior of mass concrete. Summary of Testing Program The laboratory testing program for the prediction of the behavior of mass concrete should include the following tests: 1. Isothermal calorimetry for the mea surement of heat generation rate 2. Compressive strength per ASTM 1074 or isothermal calorimetry testing for the determination of Ea 3. Compressive strength of concrete cylinders 4. Splitting tensile strength of concrete cylinders 5. Compressive modulus of elasticity at 30% of ultimate strength (for the determination of tensile modulus of elasticity) 6. Coefficient of Thermal Expansion 7. Thermal diffusivity and Specific heat capacity

PAGE 164

164 CHAPTER 11 CONCLUSIONS AND RECOMMENDATIONS FOR FUTU RE WORK Conclusions The following conclusions were drawn from this study as follows: 1. Isothermal conduction calorimetry is a viable testing method for the measurement of heat of hydration of cementitious materials. 2. The calculation of Ea using isothermal calorimetry testing provides larger valu es when compared with the calculation of Ea using the standardized test method per ASTM 1074. 3. Semi adiabatic calorimetry does not measure as much energy evolved from a cementitious system when compared with isothermal or solution calorimetry. Isothermal ca lorimetry measured the largest amount of heat evolved from a cementitious system and therefore is the most conservative method currently available for the measu rement of heat of hydration of cementitious systems. 4. The measured heat of hydration of Portland cement cured at higher temperatures is lower than Portland cement cured at lower temperatures for the same equivalent age. 5. The measured heat of hydration of cementitious systems with large replacements of Portland cement with blast furnace slag or fly ash cured at higher temperatures is higher than the same cementitious systems cured at lower temperatures. 6. The Nurse Saul maturity method (per ASTM 1074) does not accurately characterize physical and strength parameters bases for concrete with different time temperature histories. 7. The equivalent age method (per ASTM 1074) accurately models the evolution of physical and strength parameters with different time temperature histories for concrete made with Portland cement as the only cementitious material. 8. Neith er the Nurse -Saul nor the equivalent age method accurately model the evolution of strength and physical properties for cement concrete of cementitious systems with large replacements of Portland cement with blast furnace slag or fly ash. 9. The tensile and compressive moduli of elasticity of concrete are approximately equal in value for stresses below 30% of ultimate strength. 10. The thermal properties of concrete including the coefficient of thermal expansion, diffusivity, and specific heat capacity which incorporates Florida limerock are different than the published values of concrete using other limestones. Recommendations The following recommendations are based on the findings and conclusions from this study:

PAGE 165

165 1. The laboratory testing program developed in t his study should be implemented for the use in finite element modeling of mass concrete structural elements. 2. Isothermal conduction calorimetry should be standardized for the measurement of the heat of hydration of cementitious materials. 3. The standard pra ctice for estimating c oncrete strength by the maturity m ethod (ASTM C 1074) should be revised to include isothermal calorimetry testing for the calculation of apparent activation energy which can then be used for the calculation of equivalent age per equat ion 6.3 4. Accurate maturity/equivalent age models should be developed for cementitious systems which have large replacements of Portland cement with blast furnace slag or fly ash. 5. The standard test method for the measurement of the compressive modulus of el asticity (ASTM C469) should be revised to include comparative statements between tensile and compressive modulus of elasticity at stress values 30% of ultimate strength. 6. Research should be performed which involves the measurement of specific heat capacity of different concrete and constituent materials of concrete.

PAGE 166

166 APPENDIX A METHOD OF TESTING FOR MEASURING THE HEAT OF HYDRATION OF HYDRAULIC CEMENT USING ISOTHERMAL CONDUC TIVE CALORIMETRY This test method was used to determine the rate of heat of hydrati on from hardened cement itious materials by isothermal conduction calorimetry This procedure was written in an effort to conform to the procedure currently being developed by ASTM committee C01 for measurement of heat evolution of cementitious materials at early ages using the internal mixing method. Terminology Isothermal conduction calorimeter A colorimeter that measures heat flow form a sample maintained at a constant temperature by intimate thermal contact with a constant termperature heat sink. Th ermal power Heat production rate measured in watts (W) or joules per second (J/s), and is the property measured by the calorimeter. Heat Heat is the time intergral of thermal power measured in Joules (J). Baseline The signal from the calorimeter when a sample of approximately the same mass and thermal properties as a cement sample, but which is not undergoing any exothermic or endothermic reactions. Reference cell A heat -flow cell that is dedicated to outputting power from a sample that is generatin g no heat. The purpose of the reference cell is to correct for certain errors caused by drift and other systematic errors that can occur in heat -flow measuring equipment. Summary of Test Method An isothermal heat conduction calorimeter consists of a cons tant temperature heat sink to which two heat -flow sensor and sample holders are attached with good thermal conductivity. One heat flow sensor and sample holder contains the sample of interest. The other heat -flow sensor is a reference cell containing a bla nk sample (in this case glass beads) that evolves no heat. The heat of hydration released by the reacting cementitious sample is passed across the sensor and into the heat sink. The output from the calorimeter is the difference in heat flow (thermal power) the sample cell and the reference cell. The heat -flow sensor actually senses a small temperature gradient that develop from the sample side to the heat -sink side, however the heat is removed from the hydrating sample fast enough that, for practical purpo ses, the sample is at a constant temperature (isothermal). The output from the heat -flow sensor is an mV signal that is proportional to the thermal power from the sample. This output must be calibrated to a known thermal power. In this method, this is ac complished by measurements on a sample that emits a constant and known power level. The integral of the thermal power over the time of the test is the heat of hydration. Significance and Use This method is suitable for determining the total heat of hydrat ion of hydraulic cement at constant temperature at ages up to 7 days to confirm specification compliance. It gives test results equivalent to ASTM C 186 up to 7 days of age.

PAGE 167

167 Apparatus Balance accurate to 0.001 g. Volumetric dispenser device for measur ing volume or mass of water, accurate to 0.1 mL. This could be a syringe, pipette, weighting, etc. Sample holder holds the cement paste and provides intimate contact with the colorimeter heat sensing device and prevents evaporation of Mixing water. Resistance heater frabricated from material with similar heat capacity and shape as the test sample, but containing a resistor connected to a constant -voltage power supply such that a stable output of 0.01 0.0002 watt can be generated. Blank specimen fabr icated from material with similar heat capacity and shape as the test sample. Multimeter an instrument for measuring DC voltage and resistance to an accuracy of 1% over the range of measurements required for calibration and execution of the test (Note 2) This instrument is only required if the user is not following instrument specific calibration procedures. Power supply A constant voltage DC power supply with a voltage output range of at least 0 to 10 volts and a power rating of at least 0.25 watts. Calorimeter The schematic design of a calorimeter is given in Figure A 1. It consists a sample holder for the test and a reference sample, each thermally connected to heat flow sensors, which are thermally connected to a constant temperature heat sink. The minimum sensitivity for measuring heat output is 100 W. Figure A1. A schematic drawing of a heat conduction calorimeter The baseline (the power output when no heat is being generated by the sample, U0 in the calibration sequence) shall exhibit a low random noise level and be stable against drift. The rate of change of the baseline measured during a time period of three days shall be less than or equal to 20 W per gram of sample, per hour and a baseline random noise level of less than or equal to 10 W per gram sample. In practive the baseline is measure f or three days and a straight line is fitted to the data using a linear least squares procedure. The long term drift is then the slope in the line and the baseline noise level is the standard deviation around this regression line. thermostat heat sinksample referenceheat flow sensors

PAGE 168

168 Data acquisition equipmen t The data acquisition equipment used for this experiment was capable of performing continuous logging of the calorimeter output measurement at a time interval of 60 seconds. It is useful, for purposes of reducing amount of data, to have the flexibility to adjust the reading interval up to 30 minutes when power output from the sample is low. Instrument Calibration and Operating Parameters The objective of electrical calibration is to calculate a calibration constant for each individual twin calorimetric channel. The calibration constant was entered into the PicoLog software as a factor to correct the amplifier output to read the experimental results directly in mW. Each channel has a permanent precision calibration heater on side A, the sample side. Side B, used for the inert reference, does not need to be calibrated. The heater has a resistance of 100 0.1 and a very low temperature coefficient. The eight calibration heaters are connected in series together with a reference calibration resistor, also 100 0.1 To provide the calibration power, an internal power supply was turned on through a toggle switch, marked Calibration Power on the front panel of TAM Air. The calibration power applied was in the range of 35 mW, independent on the measuring r ange. During the calibration a digital voltmeter is connected to the two sockets marked Voltmeter. This voltage measurement w as used as part of the calculation of the calibration constant. Calibration was performed with empty calorimetric channels. On ce a stable baseline has been achieved, a known voltage is applied over the calibration heater by switching the toggle switch. A stable signal, as shown below, indicates that the power input was leaving the measurin g area at the same rate as it was applied This is called a steady state. Steady state calibration is simple to evaluate and does not require any form of integration as shown in Figure A -2. Figure A 2. Steady state calibration plot The calibration was performed regularly, prior to performin g each isothermal temperature for example, once calibration was performed at 15 C and 23 C etc. Each channel was independently calibrated.

PAGE 169

169 Testing Procedure The calorimeter equipment and data acquisition unit was turned on. It was determined that the calorimeter was at temperature equilibrium by verifying that the baseline (0.0V) was stable over a period of a few minutes. The temperature of the heat sink during the test was 15.0, 23.0 1.0 C. Cement Specimens o 4 grams of cementitious material was weig hed the mass recorded to the nearest 0.001 grams, and was placed in the calorimeter cell. For each of the Mixes, 2.000 grams of was used to provide a consistent water -cement ratio of at least 0.50 for each paste. After the water and cement are weighted, the cell was placed in the calorimeter as shown in Figure A-3. o Allow any change in calorimeter output caused by this process to return to the baseline level. Typically, a 24 hour interval is necessary for a return to the baseline level. Mortar Specimens o 4 g rams of cementitious material was weighed the mass recorded to the nearest 0.001 grams, and was placed in the calorimeter cell. For each of the Mixes, 2.000 grams of was used to provide a consistent water cement ratio of at least 0.50 for each paste. o 6 grams of silica sand was weighted to the nearest 0.001g and added to the calorimeter cell o After the water and cement are weighted, the cell was placed in the calorimeter same manner as the procedure used for the cement specimens noted above. After the baseline level was reached, the data collection was started and the water was injected into the cementitious materials to form a uniform paste. Cementious paste specimens were Mixed for 1 minute each. Mortar specimens were Mixed for 2 minutes each. Dat a was collected on an interval of 1 minute for the duration of the test (72 336 hours). Calculation or Interpretation of Results The purpose of the evaluation is to calculate the heat produced during the first 4 14 days of hydration. The evaluation met hod consists of the following steps: Remove the baseline: (A 1) Here Uraw is the signal from the calorimeter and Ubl is the measured baseline of the calorimeter. Apply the calibration coefficients (A and B) and divide by the mass of cement ( mc) in the sample to get the specific thermal power P : (A2)

PAGE 170

170 where P (J/s/g) is the power output per gram of dry cement, A and B are the calibration coefficients determined during the calibration process, U(t) is the voltage output at each data collection point, and mc is the mass of dry cement used in the test. The total heat of hydration of the sample is calculated by integrating the power/g versus time data over the time interval of the test ( ti to te): (A3) Where, Q (J/gcement) is the heat produced from the sample, ti is the time the sample was charged into the calorimeter, and te is the time of the end of the measurement as calculated from the time of Mixing cement and water. Operationally, the integration is executed by averaging the power output from two consecutive readings and multiplying by the time interval of the reading, giving an output for each time increment in units of J/g. The heat so calculated in each time increment is then summed over the duration of the test, as in the following equation. (A4) Where P(ti) is the power output at time ti, and P(ti+1) is the power output at the next time interval (ti+1). In this method (internally Mixed procedure), ti is taken as zero when water is added to the cement.

PAGE 171

171 Figure A 3. Cutaway of one of the 8 calorimetric channels showing the twin configuration

PAGE 172

172 APPENDIX B RAW DATA MORTAR CUBES AND APPARENT ACTIVATION ENE RGY CALCULATIONS Table A -1 Raw d ata sheet Mix 1 cubes 100% Portland c ement Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 2 Day 1 4260 1065 9:58 AM 2 4360 1090 977 9:59 AM 49:10 3 3100 775 10:01 AM 4 Day 1 12630 3158 5:41 PM 2 12120 3030 3108 5:42 PM 129:03 3 12540 3135 5:44 PM 7 Day 1 15900 3975 11:39 AM 2 15360 3840 3856 11:42 AM 171:02 3 15010 3753 11:44 AM 15 Day 1 21800 5450 11:47 AM 2 21800 5450 5442 11:49 AM 363:10 3 21700 5425 11:51 AM 28 Day 1 28400 7100 1:39 PM 2 28700 7175 7025 1:42 PM 821:02 3 27200 6800 1:44 PM 56 Day 1 27900 6975 10:00 AM 2 32300 8075 7742 10:03 AM 1657:24 3 32700 8175 10:07 AM Temperature = 8C Time of Mix = 12/9/2008 8:49 Table A -2. Raw d ata s heet Mix 1 c ubes 100% Portland c ement Age Specimen # Ultimate Load (lb) Strength (p si) Avg. Strength (psi) Time of Test Age (hours) 1 Day 1 6730 1712 10:10 AM 2 6660 1695 1699 10:12 AM 24:57 3 6640 1690 10:15 AM 2 Day 1 11350 2888 10:14 AM 2 11190 2847 2864 10:16 AM 49:01 3 11230 2858 10:18 AM 3 Day 1 14400 3664 5:15 PM 2 13920 3542 3662 5:17 PM 80:02 3 14850 3779 5:19 PM 7 Day 1 19810 5041 11:10 AM 2 19650 5000 5069 11:12 AM 169:57 3 20300 5165 11:15 AM 28 Day 1 33950 8639 11:25 AM 2 30970 7880 8266 11:30 PM 678:15 3 32540 8280 11:35 AM Temperature = 23C Time of Mix = 6/18/2008 9:15

PAGE 173

173 Table A -3 Raw d ata s heet Mix 1 c ubes 100% Portland c ement Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 1 Day 1 6730 1712 10:10 AM 2 6660 1695 1699 10:12 AM 24:57 3 6640 1690 10:15 AM 2 Day 1 11350 2888 10:14 AM 2 11190 2847 2864 10:16 AM 49:01 3 11230 2858 10:18 AM 3 Day 1 14400 3664 5:15 PM 2 13920 3542 3662 5:17 PM 80:02 3 14850 3779 5:19 PM 7 Da y 1 19810 5041 11:10 AM 2 19650 5000 5069 11:12 AM 169:57 3 20300 5165 11:15 AM 28 Day 1 33950 8639 11:25 AM 2 30970 7880 8266 11:30 PM 678:15 3 32540 8280 11:35 AM 14 Day 1 28000 7000 3:13 PM 2 29800 7450 7300 3:16 PM 4494:01 3 29800 7450 3:19 PM Temperature = 38C Figure B -1. C ompressive s trength vs. time hyperbolic m odel for Mix 1 (ASTM C 1074)

PAGE 174

174 Figure B -2 Compressive s trength vs. t ime e xponential m odel for Mix 1 (Constant Figure B -3. Compressive st rength vs. time exponential model for Mix 1 (constant u

PAGE 175

175 Table A -4. Raw d ata s heet Mix 2 c ubes 50% Portland c ement 50% s lag Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 2 Day 1 1676 419 10:03 AM 2 1694 424 429 10:04 AM 49:15 3 1783 446 10:05 AM 4 Day 1 5540 1385 5:46 PM 2 5460 1365 1338 5:48 PM 129:10 3 5050 1263 5:53 PM 7 Day 1 6640 1660 11:31 AM 2 6800 1700 1710 11:34 AM 170:54 3 7080 1770 11:36 AM 15 Day 1 10880 2720 11:53 AM 2 11570 2893 2760 11:55 AM 363:16 3 10670 2668 11:57 AM 28 Day 1 17580 4395 1:50 PM 2 17610 4403 4300 1:52 PM 821:13 3 16410 4103 1:55 PM 56 Day 1 22800 5700 9:50 AM 2 22600 5650 5592 9:52 AM 1657:13 3 21700 5425 9:55 AM Temperature = 8C Time of Mix = 12/9/2008 8:49 Table A -5. Raw d ata s heet Mix 2 c ubes 50% Portland c ement 50% s lag Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 1 Day 1 1682 428 8:30 AM 2 1466 373 401 8:33 AM 23:28 1 Day 1 2090 532 1:30 PM 2 2100 534 534 1:35 PM 28:32 3 2110 537 1:40 PM 2 Day 1 4720 1201 9:15 AM 2 4650 1183 1167 9:17 AM 48:14 3 4390 1117 9:19 AM 3 Day 1 6960 1771 8:30 AM 2 6750 1718 1751 8:32 AM 71:29 3 6930 1763 8:34 AM 7 Day 1 15900 4046 8:20 AM 2 15270 3885 3997 8:27 AM 167:26 3 15960 4061 8:40 AM 28 Day 1 28400 7226 11:25 AM 2 26600 6768 7031 11:31 AM 674:27 3 27900 7099 11:35 AM Temperature = 23C Time of Mix = 7/22/2008 9:03

PAGE 176

176 Table A -6. Raw data s heet Mix 2 c ubes 50% Portland cement 50% s lag Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 12 Hour 1 1508 377 5:38 PM 2 1545 386 383 5:40 PM 9:01 3 1540 385 5:42 PM 24 Hour 1 6830 1708 8:34 AM 2 6920 1730 1714 8:35 AM 23:56 3 6820 1705 8:36 AM 36 Hour 1 8070 2018 6:15 PM 2 8960 2240 2228 6:19 PM 33:39 3 9710 2428 6:20 PM 3 Day 1 16510 4128 3:41 PM 2 16790 4198 4166 3:44 PM 79:04 3 16690 4173 3:44 PM 7 Day 1 25300 6325 10:49 AM 2 24900 6225 6283 10:52 AM 170:12 3 25200 6300 10:54 AM 14 Day 1 32800 8200 3:24 PM 2 32200 8050 8233 3:27 PM 318:48 3 33800 8450 3:30 PM Temperature = 38C Time of Mix = 12/9/2008 8:39 Figure B -4. Compressive strength vs. time hyperbolic model for Mix 2 (ASTM C 1074)

PAGE 177

177 Figure B -5. Compressive strength vs. time exponential model for Mix 2 (Constant Figure B -6. Compressive st rength vs. time exponential model for Mix 2 (constant u

PAGE 178

178 Table A -7. Raw d ata s heet Mix 3 c ubes 65% Portland c ement 35% f ly ash Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 2 Day 1 2920 730 10:24 AM 2 2320 580 626 10:26 AM 49:47 3 2270 568 10:27 AM 4 Day 1 4820 1205 4:05 PM 2 4350 1088 1152 4:07 PM 103:41 3 4650 1163 4:09 PM 7 Day 1 7130 1783 10:45 AM 2 7920 1980 1853 10:46 AM 170:20 3 7180 1795 10:48 AM 15 D ay 1 11340 2835 5:49 PM 2 9680 2420 2557 5:51 PM 369:25 3 9660 2415 5:53 PM 28 Day 1 13410 3353 2:41 PM 2 13190 3298 3200 2:43 PM 822:17 3 11800 2950 2:45 PM 56 Day 1 18460 4615 10:12 AM 2 19060 4765 4717 10:15 AM 1513:48 3 19080 4770 10:17 AM Temperature = 8C Time of Mix = 8:38 AM Table A -8. Raw d ata s heet Mix 3 cubes 65% Portland c ement 35% f ly ash Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 1 Day 1 2720 692 8:29 AM 2 3030 771 737 8:30 AM 22:46 3 2940 748 8:32 AM 2 Day 1 5443 1385 9:09 AM 2 5003 1273 1362 9:11 AM 47:26 3 5612 1428 9:12 AM 3 Day 1 6681 1700 9:20 AM 2 7066 1798 1819 10:20 AM 72:36 3 7703 1960 11:20 AM 7 D ay 1 11620 2957 9:13 AM 2 10720 2728 2865 9:16 AM 127:31 3 11440 2911 9:18 AM 14 Day 1 9710 2471 9:30 AM 2 9040 2300 2533 9:35 AM 335:50 3 11110 2827 9:37 AM 28 Day 1 20200 5140 11:04 AM 2 27400 6972 6234 11:09 AM 617:25 3 25900 6590 11:14 AM Temperature = 23C Time of Mix = 9:44 AM

PAGE 179

179 Table A -9 Raw data s heet Mix 3 c ubes 65% p ortland c ement 35% f ly ash Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 12 Hour 1 3650 91 3 8:55 PM 2 3570 893 894 8:56 PM 12:30 3 3510 878 8:58 PM 24 Hour 1 5930 1483 8:29 AM 2 5570 1393 1448 8:31 AM 24:04 3 5880 1470 8:32 AM 36 Hour 1 6190 1548 7:31 PM 2 7430 1858 1684 7:33 PM 35:07 3 6590 1648 7:35 PM 3 Day 1 9810 2453 5:35 PM 2 11100 2775 2524 5:37 PM 81:10 3 9380 2345 5:38 PM 7 Day 1 17130 4283 10:25 AM 2 15990 3998 3976 10:27 AM 170:01 3 14590 3648 10:29 AM 14 Day 1 23200 5800 10:59 AM 2 25100 6275 6163 11:02 AM 338: 36 3 24200 6050 11:05 AM Temperature = 38C Time of Mix = 8:26 AM Figure B -7. Compressive strength vs. time hyperbolic model Mix 3 (ASTM C 1074)

PAGE 180

180 Figure B -8. Compressive strength vs. time exponential model for Mix 3 (Constant Figure B-9 Compressive strength vs. time exponential model for Mix 3 (constant u

PAGE 181

181 Table A -10. r aw data sheet Mix 4 cubes 50%-30%-20% blend (Cement s lag f ly ash) Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (ho urs) 2 Day 1 1473 368 10:28 AM 2 1535 384 380 10:29 AM 49:24 3 1555 389 10:30 AM 4 Day 1 4150 1038 4:12 PM 2 3930 983 968 5:12 PM 104:07 3 3530 883 6:12 PM 7 Day 1 5770 1443 10:40 AM 2 5860 1465 1442 10:41 AM 169:36 3 5670 1418 10:43 AM 15 Day 1 8870 2218 2:40 PM 2 8850 2213 2283 2:42 PM 365:37 3 9680 2420 2:44 PM 28 Day 1 14030 3508 2:47 PM 2 14200 3550 3505 2:49 PM 821:44 3 13830 3458 2:51 PM 56 Day 1 15380 3845 10:21 AM 2 15200 3800 3818 10:23 AM 1513:18 3 15240 3810 10:25 AM Temperature = 8C Time of Mix = 9:05:00 AM Table A -11. Raw data sheet Mix 4 cubes 50%-30%-20% blend (Cement s lag f ly ash) Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Ti me of Test Age (hours) 1 Day 1 2180 555 9:13 AM 2 1950 496 528 9:15 AM 23:59 3 2100 534 9:16 AM 2 Day 1 4360 1109 9:09 AM 2 4560 1160 1137 9:11 AM 47:55 3 4480 1140 9:12 AM 3 Day 1 6650 1692 9:20 AM 2 6940 1766 1740 9:23 AM 72:08 3 6930 1763 9:26 AM 7 Day 1 13620 3466 9:21 AM 2 12110 3081 3228 9:24 AM 168:09 3 12330 3137 9:27 AM 14 Day 1 17310 4405 9:30 AM 2 17420 4433 4416 9:33 AM 336:18 3 17340 4412 9:37 AM 28 Day 1 31000 7888 11:04 AM 2 30200 7684 7651 11:08 AM 673:52 3 29000 7379 11:11 AM Temperature = 23C Time of Mix = 9:15:00 AM

PAGE 182

182 Table A -12. Raw data sheet Mix 4 cubes 50%-30%-20% blend (Cement s lag f ly ash) Age Specimen # Ultimate Load (lb) Strength (psi) Avg. Strength (psi) Time of Test Age (hours) 12 Hour 1 2580 645 9:00 PM 2 2600 650 640 9:03 PM 12:09 3 2500 625 9:05 PM 24 Hour 1 5420 1355 8:52 AM 2 6110 1528 1453 8:53 AM 24:00 3 5910 1478 8:54 AM 36 Hour 1 7420 1855 9:07 PM 2 775 0 1938 1901 9:08 PM 36:15 3 7640 1910 9:10 PM 3 Day 1 12730 3183 5:40 PM 2 12050 3013 3169 5:42 PM 80:49 3 13250 3313 5:44 PM 7 Day 1 21700 5425 10:33 AM 2 22800 5700 5583 10:35 AM 169:42 3 22500 5625 10:38 AM 14 Day 1 2 9700 7425 11:10 AM 2 30700 7675 7292 11:13 AM 338:20 3 27100 6775 11:17 AM Temperature = 38C Time of Mix = 8:53:00 AM Figure B 10. Compressive strength vs. time hyperbolic model Mix 4 (ASTM C 1074)

PAGE 183

183 Figure B -11. Compressive strength vs. time exponential model for Mix 4 (Constant Figure B -12 Compressive strength vs. time exponential m odel for Mix 4 (c onstant u

PAGE 184

184 Figure B -13 Summary of t ime of set d ata

PAGE 185

185 APPENDIX C ISOTHERMAL CALORIMET RY RAW DATA AND ANAL YSIS Calorimetry Data of Cementitious Specimens The data presen ted in this appendix are test results from isothermal calorimetry testing. Each linear plot is an average plot of four individual specimens. The figures in this appendix reflect various analysis techniques derived from the Power vs. Test duration curves. M aturity calculations were performed using the Nurse -Saul technique using the datum temperatures established per Table C 1. The apparent activation energy used to calculate equivalent age for each respective graph was obtained per the ASTM C 1074 or hyperbo lic calculation per the testing as shown in Table C 1. The figures which display power vs. equivalent age include the first 15 minutes of data (in logarithmic scale) where the other figures do not. The reason for this display technique is the data appears as a vertical line at very early ages on a when viewed on a typical time scale. Typically the data during the first 30 minutes of mixing results in larger readings of power. Each specimen contained 4 grams of cementitious material and 2 grams of water pri or to mixing via the internal mixing method. Table C -1. Mixture c omponents, r espective a pparent activation e nergy and datum t emperatures Mix Name Mix Components Ea Datum Temperature (C) Mix 1 100 % Portland Cement 3 5642 0.1 Mix 2 50% Portland 50% Sla g 33 688 0.4 Mix 3 65% Portland 35% Fly-Ash 25757 0.45 Mix 4 50 % 30 % 20 % Blend Portland Slag Fly ash 30013 9.63

PAGE 186

186 Figure C -1 Power vs. t ime for Mix 1 Figure C -2 Energy vs. time for Mix 1

PAGE 187

187 Figure C -3. Power vs. maturity for Mix 1 Figure C -4. Energy vs. maturity for Mix 1

PAGE 188

188 Figure C 5. Power vs. equivalent age for Mix 1 Figure C 6. Energy vs. equivalent age for Mix 1

PAGE 189

189 Figure C 7. Power vs. time for Mix 2 Figure C 8. Energy vs. time for Mix 2

PAGE 190

190 Figure C 9. Power vs. maturity for Mi x 2 Figure C 10. Energy vs. maturity for Mix 2

PAGE 191

191 Figure C 11. Power vs. equivalent age for Mix 2 Figure C 12. Energy vs. equivalent age for Mix 2

PAGE 192

192 Figure C 13. Power vs. time for Mix 3 Figure C 14. Energy vs. time for Mix 3

PAGE 193

193 Figure C 15. Powe r vs. maturity for Mix 3 Figure C 16. Energy vs. maturity for Mix 3

PAGE 194

194 Figure C 17. Power vs. equivalent age for Mix 3 Figure C 18. Energy vs. equivalent age for Mix 4

PAGE 195

195 Figure C 19. Power vs. time for Mix 4 Figure C 20. Power vs. energy for Mix 4

PAGE 196

196 Figure C 21. Power vs. maturity for Mix 3 Figure C 22. Energy vs. maturity for Mix 4

PAGE 197

197 Figure C 23. Power vs. equivalent age for Mix 4 Figure C 24. Energy vs. equivalent age for Mix 4

PAGE 198

198 Calorimetry Data of Mortar Specimens The data presented in t his appendix are test results from isothermal calorimetry testing for the specimens which contained mortar. Maturity calculations were performed using the Nurse Saul technique using the datum temperatures established per Table C 1. The apparent activation energy used to calculate equivalent age for each respective graph was obtained per the ASTM C 1074 or hyperbolic calculation per the testing as shown in Table C 1. Each specimen contained 4 grams of cementitious material, 2 grams of water and 6 grams of Ot tawa sand per the standard specification ASTM C 778 prior to mixing via the internal mixing method. Figure C 25. Power vs. time for Mortar Mix 1

PAGE 199

199 Figure C 26. Energy vs. time for Mortar Mix 1 Figure C 27. Power vs. m aturity for Mortar Mix 1

PAGE 200

200 Fig ure C -28 Power vs. m aturity for Mortar Mix 1 Figure C -29 Power vs. equivalent a ge for Mortar Mix 1

PAGE 201

201 Figure C 30. Energy vs. e quivalent a ge for Mortar Mix 1 Figure C 31. Power vs. time for Mortar Mix 3

PAGE 202

202 Figure C 32. Energy vs. t ime for Mortar Mix 3 Figure C 33. Power vs. m aturity for Mortar Mix 3

PAGE 203

203 Figure C 34. Energy vs. m aturity for Mortar Mix 3 Figure C 35. Power vs. e quivalent a ge for Mortar Mix 3

PAGE 204

204 Figure C 36. Energy vs. e quivalent a ge for Mortar Mix 3 Figure C 37. Isothermal ca lorimetry data and plot s used for the calculation of activation energy for Mix 1

PAGE 205

205 Figure C 38. Isothermal calorimetry data and plots used for the calculation of activation energy for Mix 2 Figure C 39. Isothermal calorimetry data and plots used for the calculation of activation energy for Mix 3

PAGE 206

206 Figure C 40. Isothermal calorimetry data and plots used for the calculation of activation energy for Mix 4

PAGE 207

207 APPENDIX D SEMI ADIABATIC TEMPE RATURE DATA Table D-1. Mixture d esigns for c oncrete s pecimens Material Mix 1 100% Portland Cement (lb/yd3) Mix 2 50% Portland 50% Slag (lb/yd3) Mix 3 65% Portland 35% fly ash (lb/yd 3 ) Mix 4 5030 20 blend (lb/yd 3 ) Cement 681 341 443 341 Fly ash 0 0 238 136 GGBF Slag 0 341 0 204 Water 341 341 341 341 Fine Agg 1095 1088 1036 1050 Coarse A gg 1650 1668 1660 1650 Figure D -1 Typical w ater c alibration

PAGE 208

208 Figure D -2 Semi -a diabatic temperature measurements for mix 1 (laboratory mix) Figure D -3 Semi -a diabatic temperature measurements for concrete mix 1 (lab and delivered) mixes

PAGE 209

209 Figure D -4 Semi -a diabatic temperature measurements for concrete mix 2 (lab and delivered) mixes Figure D -5 Semi -a diabatic temperature measurements for mix 3 ( delivered)

PAGE 210

210 Figure D -6 Semi -a diabatic temperature measurements for mix 4 (delivered) F igure D -7 Semi -a diabatic temperature measurements for each mix

PAGE 211

211 APPENDIX E COMBINED CALORIMETRY DATA AND ANALYSIS The data presented in this appendix are test results from isothermal, semi adiabatic and sure cure calorimetry testin g for the specimens composed of cement, mortar and fly ash respectively Maturity calculations were performed using the Nurse Saul technique using the datum temperatures established per Table C 1. The apparent activation energy used to calculate equivalent age for each respective graph was obtained per the ASTM C 1074 or hyperbolic calculation per the testing as shown in Table C 1. Each isothermal calorimetry specimen contained 4 grams of cementitious material, 2 grams of water and 6 grams of Ottawa sand per the standard specification ASTM C 778 prior to mixing via the internal mixing method. The mortar specimens tested via the semi adiabatic calorimeter were composed of cementitious material, water and Ottawa sand as per Table E -1 Table E -1 Mixture d esign s for mortar s pecimens u sed in the s emi adiabatic c alorimeter Material Mix 1 100% Portland Cement (lb/yd3) Mix 3 65% Portland 35% fly ash (lb/yd 3 ) Cement 1231 800 Fly ash 0 431 GGBF Slag 0 0 Water 615 615 Fine Agg 18 50 1 850 Coarse A gg 0 0

PAGE 212

212 Figure E -1 Comparison of the time temperature histories between the Sure Cure system and the IQ drum for Mix 1 Figure E -2 Comparison of the time temperature histories between the Sure Cure system and the IQ drum for Mix 2

PAGE 213

213 Figure E -3 Compariso n of the time temperature histories between the Sure Cure system and the IQ drum for Mix 3 Figure E -4 Comparison of the time temperature histories between the Sure Cure system and the IQ drum for Mix 4

PAGE 214

214 Figure E 5. Time -temperature histories obtained b y the Sure Cure for each mix Figure E 6. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 1.

PAGE 215

215 Figure E 7. Comparison of equivalent age energy histories between semi adiabat ic calorimetry system and isothermal calorimetry for Mix 2. Figure E 8. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 3.

PAGE 216

216 Figure E 9. Comparison of equivalent age energ y histories between semi adiabatic calorimetry system and isothermal calorimetry for Mix 4. Figure E 10. Comparison of equivalent age energy histories between semi adiabatic calorimetry system and the IQ drum for mortar Mix 1.

PAGE 217

217 Figure E 11. Compari son of equivalent age energy histories between semi adiabatic calorimetry system and the IQ drum for mortar Mix 3. Figure E 12. Comparison of semi adiabatic test and three different Sure Cure system offset temperatures for a trial mix.

PAGE 218

218 Figure E -13. Comparison of semi adiabatic test and four different Sure Cure system offset temperatures for a trial mix.

PAGE 219

219 APPENDIX F PHYSICAL DATA The data presented in this appendix are physical coefficient of thermal expansion, compressive, tensile and flexural str ength testing as well as modulus of elasticity testing. The equivalent age for each specimen was calculated regardless of the curing history of each cylindrical specimen. Two cylindrical specimens were taken from each mix and monitored for temperature. The refore, the laboratory cured specimens have an equivalent age slightly larger than the actual age. The term SC denotes the cylindrical specimen was match cured the Sure C ure system as explained in C hapter 8. Specimens subject to the Sure Cure have substantially larger equivalent ages with respect to laboratory cured specimens. The cylindrical specimens used in this research were 4 x 8 specimens, and were tested in compression, splitting tension modulus of elasticity, and for coefficient of thermal expa nsion in tested in accordance with ASTM C39, ASTM C496, ASTM C 469, and AASHTO TP 60 respectively. Table F -1. Raw d ata c ompressive s trength of c ylinders Mix 1 100% Portland c ement Age Specimen # Ultimate Strength (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Strength (psi) 1 Day 1 1480 25 28 578 1593 2 1630 1 Day SC 1 2270 25 57 623 2335 2 2400 2 Day 1 2280 48 50 1109 2157 2 2100 3 2090 2 Day SC 1 3000 48 116 1177 3035 2 3070 3 Day 1 2510 72 74 1663 2567 2 2630 3 2560 7 Day 1 3550 168 170 3881 3527 2 3460 3 3570 7 Day SC 1 4440 168 319 3949 4460 2 4480 14 Day 1 3830 336 338 7762 3983 2 4200 3 3920 28 Day 1 4680 672 674 15523 4710 2 4690 3 4760

PAGE 220

220 Figure F 1. Compressive strength vs. time for Mix 1 100% Portland cement Figure F 2. Compressive strength vs. maturity for Mix 1 100% Portland cement

PAGE 221

221 Figure F 3. Compressive strength vs. equivalent age for Mix 1 100% Portland cement Table F -2 Raw d ata compressive s trength of c ylinders Mix 2 50% Portland c ement 50% s lag Age Specimen # Ultimate Strength (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Strength (psi) 1 Day 1 300 24 25 562 283 2 280 3 270 1 Day SC 1 1390 24 45 588 1380 2 1370 2 Day 1 870 48 49 1123 860 2 840 3 870 2 Day SC 1 3160 48 103 1150 3175 2 3190 3 Day 1 1380 72 73 1685 1363 2 1240 3 1470 7 Day 1 2450 168 169 3931 2543 2 2560 3 2620 7 Day SC 1 4710 168 303 3958 4730 2 4750 14 Day 1 3310 336 337 7862 3420 2 3410 3 3540 28 Day 1 4820 672 673 15725 4513 2 4450 3 4270

PAGE 222

222 Figure F 4. Compressive strength vs. time for Mix 2 5 0% P ortland cement 50% s lag Figure F 5. Compressive strength vs. maturity for Mix 2 5 0% Portland cement 50% s lag

PAGE 223

223 Figure F6. Compressive strength vs. equivalent age for Mix 2 50% Portland 50% s lag Table F -3. Raw d ata c ompressive s trength of cy linders Mix 2 6 5% Portland c ement 35% f ly ash Age Specimen # Ultimate Strength (psi) Time (hours) Equivalent Age (hours) Maturity (C hr.) Avg. Strength (psi) 1 Day 1 740 25 27 586 720 2 730 3 690 1 Day SC 1 1220 25 45 612 1225 2 1230 2 Day 1 1080 48 51 1126 1057 2 990 3 1100 2 Day SC 1 1750 48 89 1175 1765 2 1780 3 Day 1 1330 72 75 1688 1390 2 1440 3 1400 7 Day 1 1820 168 171 3940 1807 2 1770 3 1830 7 Day SC 1 2770 168 270 3989 2800 2 2830 14 Day 1 2310 336 339 7879 2297 2 2530 3 2050 28 Day 1 4680 672 675 15758 4710 2 4690 3 4760

PAGE 224

224 Figure F 7. Compressive strength vs. time for Mix 3 65 % Portland cement 35% f ly ash Figure F 8. Comp ressive strength vs. maturity for Mix 3 65 % Portland cement 35% fly ash

PAGE 225

225 Figure F 9. Compressive strength vs. equivalent age for Mix 3 65% Portland cement 35% f ly ash Table F -4. Raw d ata c ompressive s trength of c ylinders Mix 2 50% Portland cem ent 30% s lag 20% fly ash Age Specimen # Ultimate Strength (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Strength (psi) 1 Day 1 260 25 25 816 297 2 340 3 290 1 Day SC 1 900 25 35 846 810 2 720 2 Day 1 86 0 48 49 1566 867 2 870 3 870 2 Day SC 1 1920 48 73 1629 1780 2 1640 3 Day 1 1280 72 73 2349 1213 2 1170 3 1190 7 Day 1 2360 168 169 5482 2257 2 2220 3 2190 7 Day SC 1 3700 168 233 5545 3765 2 3830 14 D ay 1 3430 336 337 10964 3170 2 2860 3 3220 28 Day 1 3940 672 673 21927 4077 2 4130 3 4160

PAGE 226

226 Figure F 10. Compressive strength vs. time for Mix 4 50% Portland 30% slag 20% f ly ash Figure F 11. Compressive strength vs. m aturity for Mix 4 50% Portland 30% slag 20% f ly ash

PAGE 227

227 Figure F 12. Compressive strength vs. equivalent age for Mix 4 50% Portland 30% slag 20% f ly ash Figure F 13. Compressive strength vs age summary for each m ix

PAGE 228

228 Figure F -14 Compressive strength vs equivalent age summary for each mix Tabl e F -5 Raw d ata s plitting t ensile s trength of c ylinders Mix 1 100% Portland c ement Age Specimen # Ultimate Strength (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Strength (psi) 1 Day 1 137 25 28 578 181 2 197 3 207 1 Day SC 1 224 25 57 623 222 2 221 2 Day 1 252 48 50 1109 241 2 233 3 239 2 Day SC 1 359 48 116 1177 340 2 320 3 Day 1 222 72 74 1663 280 2 351 3 268 7 Day 1 321 168 170 3881 320 2 307 3 332 7 Day SC 1 395 168 319 3949 410 2 425 14 Day 1 493 336 338 7762 431 2 393 3 407 28 Day 1 466 672 674 15523 479 2 520 3 451

PAGE 229

229 Figure F -15. Tensile strength vs. time for Mi x 1 100% Portland cement Figure F -16 Tensile strength vs. maturity for Mix 1 100% Portland cement

PAGE 230

230 Figure F -17 Tensile strength vs. equivalent age for Mix 1 100% Portland cement Table F -6. Raw d ata s plitting t ensile s trength of c ylinders Mix 2 50% P ortland c ement 50% s lag Age Specimen # Ultimate Strength (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Strength (psi) 1 Day 1 34.7 25 25 562 37 2 39.3 3 36.5 1 Day SC 1 174 25 46 588 163 2 152 2 Day 1 107 48 49 1123 117 2 137 3 108 2 Day SC 1 367 48 107 1150 365 2 364 3 Day 1 178 72 73 1685 179 2 173 3 188 7 Day 1 319 168 169 3931 335 2 201 3 257 7 Day SC 1 489 168 311 3958 449 2 410 14 Day 1 381 336 337 7862 356 2 389 3 299 28 Day 1 399 672 673 15725 434 2 453 3 450

PAGE 231

231 Figure F -18. Tensile strength vs. time for Mix 2 5 0% Portland cement 50% s lag Figure F -19 Tensile strength vs. maturity for Mix 2 5 0% Por tland cement 50% s lag

PAGE 232

232 Figure F 20. Tensile strength vs. equivalent age for Mix 2 5 0% Portland cement 50% s lag Table F -7. Raw data s plitting t ensile strength of c ylinders Mix 2 6 5% Portland c ement 35% f ly ash Age Specimen # Ultimate Strength (p si) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Strength (psi) 1 Day 1 99 25 26 563 97 2 91 3 102 1 Day SC 1 144 25 38 612 147 2 150 2 Day 1 131 48 50 1126 135 2 139 3 134 2 Day SC 1 194 48 76 1175 190 2 185 3 Day 1 183 72 73 1688 185 2 184 3 186 7 Day 1 259 168 169 3940 236 2 211 3 238 7 Day SC 1 338 168 238 3989 310 2 282 14 Day 1 288 336 337 7879 277 2 267 3 276 28 Day 1 466 672 673 15758 479 2 520 3 451

PAGE 233

233 Figure F 21. Tensile strength vs. time for Mix 3 65% Portland cement 35% fly ash Figure F -22 Tensile strength vs. maturity for Mix 3 65% Portland cement 35% f ly ash

PAGE 234

234 Figure F -23 Tensile strength vs. equivalent age for Mix 3 65% Portland cement 35% fly ash Table F -8. Raw d ata s plitting tensile s trength of c ylinders Mix 4 50% Portland 30% slag 20% f ly ash Age Specimen # Ultimate Strength (psi) Time (hours) Equivalent Age (hours) Maturity (C hr.) Avg. Strength (psi) 1 Day 1 61 25 25 783 57 2 49 3 61 1 Day SC 1 127 25 34 846 139 2 151 2 Day 1 117 48 49 1566 111 2 94 3 121 2 Day SC 1 194 48 73 1629 187 2 181 3 Day 1 150 72 73 2349 147 2 153 3 138 7 Day 1 280 168 169 5482 264 2 247 3 265 7 Day SC 1 431 168 231 5545 350 2 270 14 Day 1 275 336 337 10964 319 2 294 3 387 28 Day 1 327 672 673 21927 414 2 477 3 438

PAGE 235

235 Figure F -24. Tensile strength vs. time for Mix 4 50% Portland 30% slag 20% f ly ash Figure F -25 Tensile strength vs. maturity for Mix 4 50% Portland 30% slag 20% f ly ash

PAGE 236

2 36 Figure F -26 Splitting tensile strength vs. equivalent age for Mix 4 50% Portland 30% slag 20% f ly ash Figure F -27 Compressive strength vs age summary for each m ix

PAGE 237

237 Figure F -28 Tensile strength vs equivalent a ge summary for each m ix Table F -9. Raw d ata c ompressive m odulus of e lasticity of c ylinders Mix 1 100% Portland c ement Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 1.94E+06 25 25 783 1.95E+06 2 1.95E+06 3 1.96E+06 1 Day SC 1 2.50E+06 25 34 846 2.50E+06 2 Day 1 2.49E+06 48 49 1566 2.45E+06 2 2.43E+06 3 2.41E+06 2 Day SC 1 2.97E+06 48 73 1629 2.97E+06 3 Day 1 2.59E+06 72 73 2349 2.62E+06 2 2.55E+06 3 2.73E+06 7 Day 1 2.88E+06 168 169 5482 2.93E+06 2 2.95E+06 3 2.97E+06 7 Day SC 1 3.22E+06 168 231 5545 3.22E+06 14 Day 1 3.21E+06 336 337 10964 3.23E+06 2 3.27E+06 3 3.21E+06 28 Day 1 3.50E+06 672 673 21927 3.66E+06 1 3.71E+06 1 3.76E+06

PAGE 238

238 Figure F -29. Compressive modulus of elasticity vs. time for Mix 1 100% P ortland cement Figure F30. Compressive modulus of elasticity vs. maturity for Mix 1 100% Portland cement

PAGE 239

239 Figure F31. Compressive modulus of elasticity vs. equivalent age for Mix 1100% Portland cement Tabl e F 10. Raw d ata m odulus of e lasticity of c ylinders Mix 2 50% Portland c ement 50% s lag Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 1.43E+06 25 25 562 1.25.E+06 2 1.25E+06 3 1.07E +06 1 Day SC 1 3.00E+06 25 46 588 3.00E+06 2 Day 1 1.90E+06 48 49 1123 1.91.E+06 2 2.15E+06 3 1.67E+06 2 Day SC 1 2.90E+06 48 107 1150 2.90E+06 3 Day 1 2.29E+06 72 73 1685 2.16.E+06 2 2.09E+06 3 2.11E+06 7 Day 1 2.86E+06 168 169 3931 2.82.E+06 2 2.73E+06 3 2.86E+06 7 Day SC 1 3.75E+06 168 311 3958 3.75E+06 14 Day 1 3.08E+06 336 337 7862 3.12.E+06 2 3.05E+06 3 3.21E+06 28 Day 1 3.61E+06 672 673 15725 3.61.E+06 2 3.39E+06 3 3.84E+06

PAGE 240

240 Figure F -32 Compressive modulus of elasticity vs. age for Mix 2 50% Portland cement 50% s lag Figure F -33 Compressive modulus of maturity vs. age for Mix 2 50% Portland cement 50% s lag

PAGE 241

241 Figure F-34. Compressive modulus of elasticity vs. equi valent age for Mix 2 50% Portland cement 50% s lag Tabl e F 11. Raw d ata m odulus of e lasticity of c ylinders Mix 2 65% Portland c ement 35% f ly ash Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 1.25E+06 25 26 563 1.42.E+06 2 1.49E+06 3 1.51E+06 1 Day SC 1 1.39E+06 25 38 612 1.39.E+06 2 Day 1 1.99E+06 48 50 1126 2.03.E+06 2 2.02E+06 3 2.09E+06 2 Day SC 1 2.26E+06 48 76 1175 2.26.E+06 3 Day 1 2.00E+06 72 73 1688 2.10.E+06 2 1.98E+06 3 2.32E+06 7 Day 1 2.41E+06 168 169 3940 2.40.E+06 2 2.54E+06 3 2.26E+06 7 Day SC 1 2.87E+06 168 238 3989 2.87.E+06 14 Day 1 2.76E+06 336 337 7879 2.75.E+06 2 2.73E +06 3 2.77E+06 28 Day 1 3.50E+06 672 673 15758 3.66.E+06 2 3.71E+06 3 3.76E+06

PAGE 242

242 Figure F -35 Compressive modulus of elasticity vs. age for Mix 3 65% Portland cement 35% f ly ash Figure F -36 Compressive modulus of maturity vs. age for Mix 3 65% Portland cement 35% f ly ash

PAGE 243

243 Figure F -37 Compressive modulus of elasticity vs. equivalent age equivalent age for M ix 3 65% Portland cement 35% f ly ash Table F 12. Raw d ata m odulus of e lasticity of c ylinders Mix 4 50% P ortlan d cement 30% slag 20% f ly ash Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 1.11E+06 25 25 783 1.12E+06 2 1.21E+06 3 1.04E+06 1 Day SC 1 1 .93E+06 25 34 846 1.93E+06 2 Day 1 1.71E+06 48 49 1566 1.59E+06 2 1.47E+06 3 1.59E+06 2 Day SC 1 2.22E+06 48 73 1629 2.22E+06 3 Day 1 2.45E+06 72 73 2349 2.58E+06 2 2.55E+06 3 2.73E+06 7 Day 1 2.47E+06 168 169 5482 2.45E+06 2 2.44E+06 3 2.45E+06 7 Day SC 1 3.28E+06 168 231 5545 3.28E+06 14 Day 1 2.82E+06 336 337 10964 2.91E+06 2 2.86E+06 3 3.04E+06 28 Day 1 3.28E+06 672 673 21927 3.39E+06 2 3.42E+06 3 3.47E+06

PAGE 244

244 Figure F -38 Compressi ve modulus of elasticity vs. age for Mix 4 50% Portland cement 30% slag 20% f ly ash Figure F -39 Compressive modulus of elasticity vs. maturity for Mix 4 50% Portland cement 30% slag 20% f ly ash

PAGE 245

245 Figure F40. Compressive modulus of elastic ity vs. equivalent age for Mix 4 50% Portland cement 30% slag 20% f ly ash Figure F 41. Compressive modulus of elasticity vs. age for each m ix

PAGE 246

246 Figure F -42 Compressive modulus of elasticity vs. age for each m ix Table F 13. Raw data t ensile m odulus of e lasticity of c ylinders Mix 1 100% Portland c ement Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 2.34E+06 25 25 783 2.33E+06 2 2.34E+06 3 2.32E+06 2 Day 1 2.61E+06 48 49 1566 2.46E+06 2 2.46E+06 3 2.31E+06 3 Day 1 2.64E+06 72 73 2349 2.54E+06 2 2.47E+06 3 2.53E+06 7 Day 1 2.76E+06 168 169 5482 2.79E+06 2 2.81E+06 3 2.80E+06 14 Day 1 2.83E+06 336 337 10964 2.99E+06 2 3.19E+06 3 2.95E+06

PAGE 247

247 Figure F -43. Tensile modulus of elasticity and compressive modulus of elasticit y vs. equivalent age for Mix 1 100% Portland cement Table F 14. Raw d ata t ensile m odulus of e lasticity of c ylinders Mix 2 50% Portland cement 50% s lag Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 1.25E+06 25 25 562 1.22E+06 2 1.21E+06 3 1.20E+06 2 Da y 1 1.84E+06 48 49 1123 1.93E+06 2 2.02E+06 3 No Data 3 Day 1 2.10E+06 72 73 1685 2.15E+06 2 2.08E+06 3 2.27E+06 7 Day 1 2.58E+06 168 169 3931 2.56E+06 2 2.50E+06 3 2.61E+06 14 Day 1 2.81E+06 336 337 7862 2.92E+06 2 2.99E+06 3 2.95E+06 28 Day 1 3.30E+06 672 673 15725 3.33E+06 2 3.35E+06 3 No Data

PAGE 248

248 Figure F -44. Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 2 50% Portland cement 50% s lag Table F -15. Raw data t ensile m odulus of e lasticity of c ylinders Mix 2 65% Portland c ement 35% f ly ash Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C -hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 1.67E+06 25 26 563 1.67E+06 2 1.67E+06 3 1.67E+06 2 Day 1 1.94E+06 48 50 1126 1.92E+06 2 1.95E+06 3 1.88E+06 3 Day 1 2.27E+06 72 73 1688 2.25E+06 2 2.28E+06 3 2.20E+06 7 Day 1 2.49E+06 168 169 3940 2.46E+06 2 2. 55E+06 3 2.35E+06 14 Day 1 2.79E+06 336 337 7879 2.78E+06 2 2.83E+06 3 2.72E+06 28 Day 1 3.06E+06 672 673 15758 3.04E+06 2 3.12E+06 3 2.95E+06

PAGE 249

249 Figure F-45. Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 3 65% Portland cement 35% f ly ash Table F 16. Raw d ata t ensile m odulus of elasticity of c ylinders Mix 4 50% Portland 30% slag 20% f ly ash Age Specimen # Modulus of Elasticity (psi) Time (hours) Equivalent Age (hours) Maturity (C hr.) Avg. Modulus of Elasticity (psi) 1 Day 1 1.18E+06 25 25 783 1.24E+06 2 1.28E+06 3 1.25E+06 2 Day 1 1.89E+06 48 49 1566 1.86E+06 2 1.80E+06 3 1.88E+06 3 Day 1 2.22E+06 72 73 2349 2.15E+06 2 2 .11E+06 3 2.12E+06 7 Day 1 2.72E+06 168 169 5482 2.61E+06 2 2.49E+06 3 2.61E+06 14 Day 1 2.87E+06 336 337 10964 2.95E+06 2 2.91E+06 3 3.07E+06 28 Day 1 3.18E+06 672 673 21927 3.11E+06 2 3.02E+06 3 3.15E+06

PAGE 250

250 Figure F -46. Tensile modulus of elasticity and compressive modulus of elasticity vs. equivalent age for Mix 4 50% Portland 30% slag 20% f ly ash Figure F-47. Tensile modulus of elasticity vs. time for each mix

PAGE 251

251 Table F 17. Raw d ata c oefficien t of thermal expansion t esting Mix 1 Mix 2 Mix 3 Mix 4 Age (days) CTE (in/in/C) Age (days) CTE (in/in/C) Age (days) CTE (in/in/C) Age (days) CTE (in/in/C) 1 9.16E 06 1 1.11E 05 1 1.04E 05 1 9.39E 06 2 9.81E 06 2 1.11E 05 2 1.08E 05 2 9.92E 06 3 No data 3 No data 3 1.09E 05 3 1.00E 05 7 9.95E 06 7 No data 7 1.10E 05 7 1.01E 05 14 9.95E 06 14 1.20E 05 14 1.12E 05 14 No data 28 9.95E 06 28 1.17E 05 28 1.16E 05 28 No data Figure F -48. Coefficient of thermal expansion vs. time for each mix

PAGE 252

252 APPENDIX G CEMENTITIOUS MATERIAL DATA AND MILL CERTIFICATIONS Figure G 1. Solution c alorimetry data for Mix 1 100% Portland c ement Figure G 2. Solution c alorimetry data for Mix 2 50% Portland c ement 50% s lag

PAGE 253

253 Figure G 3. Solution c alori metry data for Mix 3 65% Portland c ement 35% f ly ash Figure G 4. Solution c alorimetry data for Mix 3 50% Portland c ement 30% s lag 20% fly a sh

PAGE 254

254 Table G-1. X-ray f luorescence d ata for Portland c ement used in e ach m ix Compound Mix 1 Mix 2 Mix 3 Mi x 4 SiO 2 20.53 20.61 20.26 20.10 Al 2 O 3 4.85 4.94 5.02 4.96 Fe 2 O 3 4.33 4.08 4.18 4.05 CaO 63.52 63.66 63.78 63.98 MgO 0.67 0.71 0.70 0.73 SO 3 2.97 3.04 3.09 2.96 Na 2 O 0.146 0.155 0.149 0.156 K 2 O 0.39 0.39 0.41 0.45 TiO 2 0.25 0.25 0.25 0.25 P 2 O 5 0. 423 0.418 0.401 0.399 Mn 2 O 3 0.020 0.019 0.017 0.019 C3A 5.509 6.180 6.245 6.303 C3S 55.341 54.828 57.16 60.093 C4AF2 13.188 12.423 12.708 12.316 C2S 17.415 18.038 15.278 12.626 Eq. Alkalies 0.402 0.412 0.417 0.453 Table G2. X-ray diffraction data for Portland cement used in e ach m ix Phase Mix 1 Mix 2 Mix 3 Mix 4 Alite 48.81 51.76 54.20 54.91 Belite 28.94 26.05 23.69 21.92 Ferrite 13.70 12.35 13.15 13.07 Free Lime 0.40 0.17 0.38 0.30 Alum_cub 5.38 6.03 5.76 5.91 Alum_ortho 0.53 0.77 0.50 0.88 Periclase 0.17 0.25 0.00 0.20 Arcanite 0.00 0.58 0.60 0.54 Portlandite 0.83 0.49 0.36 0.78 Calcite 0.29 0.45 0.39 0.60 Quartz 0.00 0.00 0.00 0.00 Gypsum 0.00 0.00 0.00 0.00 Hemihydrate 0.95 0.85 0.85 0.78 Anhydrite 0.00 0.25 0.13 0.10

PAGE 255

255 Figure G5. Laboratory testing data of Portland cement used for Mix 1

PAGE 256

256 Figure G 6. Laboratory testing data of Portland cement used for Mix 1

PAGE 257

257 Figure G -7 Laboratory testing data of Portland cement used for Mix 2

PAGE 258

258 Figure G -8 Laboratory testing data of Portland cement used for Mix 2

PAGE 259

259 Figure G -9 Laboratory testing data of granulated blast furnace slag used for Mix 2

PAGE 260

260 Figure G 10. Laboratory testing data of granulated blast furnace slag used for Mix 2

PAGE 261

261 Figure G -11 Laboratory testing data of Portland c ement used for Mix 3

PAGE 262

262 Figure G 12. Laboratory testing data of fly ash used for Mix 3

PAGE 263

263 Figure G -13 Laboratory testing data of fly ash used for Mix 3

PAGE 264

264 Figure G -14 Mill certification data for Portland cement used in each mix

PAGE 265

265 Figure G -15 Mill ce rtification data for granulated blast furnace slag cement used in each mix

PAGE 266

266 Figure G -16 Mill certification data for fly ash cement used in each mix

PAGE 267

267 LIST OF REFERENCES AASHTO (2007) Standard Test for Coefficient of Thermal Expansion of Hydraulic Cem ent Concrete, TP 60 Washington, DC American Association of State and Highway Officials ACI Committee 207 (2005) Mass Concrete (ACI 207.1-R05), American Concrete Institute, Farmington Hills, Michigan 30 pages American Concrete Institute. Committee 207 (2005) Mass Concrete Detroit, Michigan: American Concrete Institute Ash, J.E., Hall, M.G., Langford, J.I., & Mellas, M., (1993) Estimations of Degree of Hydration of Portland Cement Pastes, Cement and Concrete Research 23(2), 399406. ASTM (2002) Standard Te st Method for Static Modulus of Elasticity and Poissons Ratio of Concrete in Compression, C 469. West Conshohocken, Pennsylvania: American Society for Testing Materials. ASTM (2003) Standard Test Method for Fineness of Portland C ement by the Turbidimeter C11596a. West Conshohocken, Pennsylvania: American Society for Testing Materials. ASTM (2004a) Standard Test Method for Steady-State Heat Flux Measurements and Thermal Transmission Properties by Means of the GuardedHot -Plate Apparatus, C17704, West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2004b) Standard Test Method for Specific Heat of Rock and Soil D 461186. West Conshohocken, Pennsylvania: American Society for Testing Materials. ASTM (2004c) Standard Practice for Estimating Concrete Strength by the Maturity Method C 107404 West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2004d) Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens C 39/C 39M, West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2005) Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2 in. or [50 -mm] Cube Specimens) C 109/C 109M West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2006a) Standard Test Method for Flexural Performance of Fiber Reinforced Concrete (Using Beam w ith Third Point Loading) C 1609/C 1609M 06 West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2006b) Standa rd Test Method for Length Change of Hardened Hydraulic Cement Mortar and Concrete, C 157/C 157M West Conshohocken, Pennsylvania: American Society for Testing Materials

PAGE 268

268 ASTM (2007a) Standard Practice for Calibrating Thin Heat Flux Transducers C 113007 We st Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2007c) Standard Practice for Making and Curing Concrete Test Specimens in the Laboratory, C 192/C 192M 07 West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2008a) Standard Practice for Creep of Concrete in Compression, C 51202 West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2008b) Standard Practice Measuring Hydration Kinetics of Hydraulic Cementitious Mixtures Using Iso thermal Calorimetry, C 167908 West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2008c) Standard Specification for Portland Cement C 15007. West Conshohocken, Pennsylvania: America n Society for Testing Materials ASTM (2008d) Standard Test Method for Density (Unit Weight), Yield, and Air Content (Gravimetric) of Concrete, 138C/138M 08M West Conshohocken, Pennsylvania: American Society for Testing Materials ASTM (2009) Standard Test Methods for Chemical Analysis of Portland Cement C 11409 West Conshohocken, Pennsylvania: American Society for Testing Materials Balendran, R.V., (1995) Estimating the Elastic Modulus of Concrete Made with Artificially Manufactured Lightweight Aggregates, Structural Survey, Vol.13,No.2, pp.1620. Barnett, S.J., Soutsos, M.N., Millard, S.G., and Bungey, J H. (2006) Strength Development of Mortars Containing Ground Granulated Blast Furnace Slag: Effect of Curing Temperature and Determination of Apparent Activation Energies Cement & Concrete Research Elsevier, 36(3) 434440 Bamforth P.B., (1984) Mass Concrete. Concrete Society Digest No. 2 Bentz, D.P. & Sutzman, P.E. (2006) Curing, Hydration, and Microstructure of Cement Paste, ACI Mat erials Journal, 103(5), 348356 Bhat ty, J.I., and Taylor, P.T., (2006) Sulfate Resistance of Concrete Using Clended Cements or Supplementary Cementitious Materials, Portland Cement Association Skokie, IL Bogue, R.H., (1947) The Chemistry of Portland Cement. New York, New York: Reinhold Pu blishing Corporation Boyd A.J. and Mindess, S., (2002) An Indirect Tension Test for Concrete, Fifth International Symposium on Cement & Concrete, Tongji University Press, Shanghai, China, I, 590594.

PAGE 269

269 Boyd, A.J and Mindess, S., (2004) The Use of Tension Testing to Investigate the Effect of W/C Ratio and Cement Type on the Resistance of Concrete to Sulfate Attack, Cement & Concrete Research Elsevier, 34(3) 373377 Bremner, T.W., Boyd, A.J., H olm, T.A. and Boyd, S.R. (1998) Indirect Tensile Testing to Evaluate the Effect of Alkali -Aggregate Reaction in Concrete, Proceedings, Structural Engineering World Wide ASCE, San Francisco, USA, CD ROM Paper T1922, 12 pp. Broda, M., Wriquin, E., Dutho it, B., (2002) Conception of an Isothermal Calorimeter for Concrete Determination of the Apparent Activation Energy Materials and Structures (35) 389394. Brown, R.W. III (1994) Estimating Heat Development in Structural Concrete Masters Thesis, Univer sity of Florida Carino, N.J ., (2004) The Maturity Method Handbook on Nondestructive Testing of Concrete 2nd Ed., Edited by Malhorta V.M., and Cari no N.J., New -York, CRC Press, 5 -1 -547 Carino, N.J. & Tank, R.C. (1992) Maturity Functions for Concrete Ma de with Various Cements and Admixtures, ACI Materials Journal, 89(2) 188196. Carino, N.J., & Lew, H.S. (1982) Temperature Effects on StrengthMaturity Relations of Mortar, ACI Journal, 80(17) 177182 Carino, N.J., Lew, H.S., & Volz, C.K., (1982) Early Ag e Temperature Effects on Concrete Strength Prediction by the Maturity Method, ACI Journal 80(10) 93101 Chini A.R., Muszynski, L.C., Acquaye, L., and Tarkhan, S., Determination of the Maximum Placement and Curing Tempratures in Mass Concrete to Avoid Durability Problems and DEF, FDOT Contract BC 35429, Florida Department of Transportation, Tallahassee February 2003 Chirdon, W.M., Aquino,W., Hover, K.C, (2006) A Method for Measuring the Transient Thermal Diffusivity in Hydrating Portland Cement Mortars Using an Oscillating Boundary Temperature, Cement and Concrete Research, 37(5) 680690 Cusson, D. and Hoogeveen, T.J, (2006) Measuring Early Age Coefficient of Thermal Expansion in High Performance Concrete, Volume Changes in Hardening Concrete: Testing and Mitigation, Edited by Jensen, M.O., Lura, P., Kolver, K., RILEM, 321 330 DAloia, L., and Chanvillard, G., (2002) Determining the apparent activation energy of concrete: Ea numerical simulations of the heat of hydration of cement Cement and Concrete Research 32(8), 12771289 DeSchutter, G., and Taewe, L (1995). Specific Heat and Thermal Diffusivity of Hardening Concrete, Magazine of Research 47(172) 203208

PAGE 270

270 Divet, L., and Pavione, A., (2002). Delayed Ettringite Formation in Massive Concrete Structure:, an Account of Some Studies of Degraded Bridges, Proceedings, of the International RILEM TC 186 ISA Workshop,RILEM, Villars, Switzerland 98126. Drimlas T.,(2004) LaboratoryTesting and Investigat ions of Delayed Ettringite Formation, Masters Thesis, University of Texas at Austin Escalante Garcia, J.I., & Sharp, J.H. (1998). Effect of Temperature on the Hydration of the Main Clinker Phases in Portland Cements: Part I, Neat Cements. Ceme nt and Conc rete Research, 28(9) 12451257 Escalante Garcia, J.I., & Sharp, J.H. (2001). The microstructure and mechanical properties of blended cements hydrated at various temperatures, Cement and Concr ete Research, 31 (5) 675702 Evju, C., (2003) Initial Hydration o f Cemenitious Systems Using a Simple Isothermal Calorimeter and Dynamic Correction, Journal of Thermal Analysis and Calorimetry 71, 829840 Ferraro, C.C., & Tia, M., (2009) Measurement of Apparent Activation Energy of Portland Cement Based Cementitious Mat erials, Presented at the International Summit of Hydration Kinetics, Quebec, CA, July 2729 Florida Department of Transportation, Standard Specifications for Road and Bridge Construction, Florida Department of Transportation, Tallahassee 2007 Gajda, J., (2 007) Mass Con crete for Buildings and Bridges Portland Cement Association Skokie, Illinois, 34 pp. Gater, R.A. (1995) Engineering Thermodynamics, Foundation Topics, Vol. 1. Ghorab, H.Y., (2002) On the Chemistry of Delayed Ettringi te Formation, Proceedings, of the International RILEM TC 186 ISA Workshop, RILEM, Villars, Switzerland, 65 81 Glasser, F.P. (2005) Thermodynamics of Cement Hydration, Materials Science of Concrete VII Edited by Young, F. & Skalny, J., Westerville, Ohio: The American Ceramic Society, 37 100. Glasstone, S., Laidler, K., & Eyring, H. (1941) The Theory of Rate Processes New York, New York: McGraw Hill Book Company Hansen, T.C. (1986). Physical Structure of Hardened Cement Paste. A Classical Approach. Mate rials and Structures, 19(114) 423436 Helal, M.A. (2001) Effect of Curing Time on the PhysicoMechanical Characteristics of the Hardened Cement Pastes Containing Limestone Cement and Concrete Research, 32, 447450.

PAGE 271

271 Higginson, E.C. (1970) The effect of ce ment fineness on concrete, Fineness of Cement, STP 473, A STM, Philadelphia, PA, pp 71 81 Hooton, R.D., Boyd, A.J., & Bhadkamkar, D. (2005) Effect of Cement Fineness and C3S Content on Properties of Concrete: A Literature Review, R&D Serial No. 2871, Portla nd Cement Association, Skokie, Illinois 16 pp. Hsu, T.T., Slate, F.O, Sturman, G.M., Winter, G, (1963) Microcracking of Plain Concrete and the Shape of the Stress Strain Curve ACI Journal, 60(2) 209224 Josephson, G.W., Sillers, F., & Runner, D.G. (1949) Iron Blast Furnace Slag: Production, Processing, Properties, and Uses Washington D.C.: United States Government Printing Office Kada -Benameur, H., Wirquin, E., Duhoit, B., (2000) Determination of the Apparent Activation Energy of Concrete by Isothermal Ca lorimetry, Cement and Concrete Research, 32(2), 301305 Kee, C.F., (1971) Relation Between Strength and Maturity of Concrete. ACI Journal, 68(21), 196203 Kesler, C.E., (1966) Strength Significance of Tests and Properties of Concrete and Concrete Making Materials ASTM STP 169, pp. 144159 Khan, A., Cook, W., & Mitchell, D. (1998). Thermal Properties and Transient Thermal Analysis of Structural Members During Hydration. ACI Materials Journal, 95(28), 293300 Kim, S.M., Won, M.C., (2004) Horizontal Cracki ng in Continuously Reinforced Concrete Pavements ACI Materials Journal, 1 01(6) 784791 Kjellsen, K.O., Detwiler, R.J., & Gjorv, O.E. (1991) Development of Microstructures in Plain Cement Pastes Hydrated at Different Temperatures. Cement and Concrete Res earch, 21 179189. Kjellsen, K.O. &, Detwiler, R.J., (1993) Later age strength Prediction by a Modified Maturity Model, ACI Materials Journal, 8 0 (3) 220227 Klein, A., Pirtz, D., & Adams, R.F. (1963) Thermal Properties of Mass Concrete During Adiabatic C uring, Symposium on Mass Concrete ACI SP 6,199218. Lane, D.S., and Ozyildrirm, H.C., (1999) Evaluation of the Potential for Internal Sulfate Attack Through Adaptation of ASTM C 342 and the Duggan Test Cement, Concrete & Aggregates ASTM, 21, (1) 43 58 Larsen, T.J (1991) The Effects of Variation in Tricalcium silicate of Portland cement on concrete properties, August 1989 Revised Feb 1991.

PAGE 272

272 Lidstrom L., and Westerberg, B. (2003) Fine Ground Cement in Concrete Properties and Pros pects, ACI Materials Journal, 100(5 ) 398406 Ma, W., Sample, D., Martin, R., & Brown, P.W., (1994) Calorimetric Study of Cement Blends Containing Fly Ash, Silica Fume, and Slag at Elevated Temperatures, Cement, Concrete, and Aggregates, CCAGPD, 16(2) 9399 Mehta, P.K., (1986) Concrete Structure, Properties, and Materials Englewood Cliffs, New Jersey: Prentice Hall, Inc. Melander, J, (2007) AASHTO and ASTM Move Toward Harmonizing for Portland Cement Standards, Concrete Infocus, A publication of the National Ready Mixed Concrete Association, Naylor LLC, Summer,40 43 Miller, F. M., and Conway, T, (2003) Use of Ground Blast Furnace Slag for the Reduction of Expansion Due to Delayed Ettringite Formation, Cement, Concrete & Aggregates ASTM, 25(1) 4358. Mindess, S., You ng, J. F., & Darwin, D. (2003) Concrete (2nd ed.). Upper Saddle River, New Jersey: Prentice Hall Mitchell L.J., (1966) Thermal Properties Significance of Tests and Properties of Concrete and Concrete -Making Materials ASTM STP 169, 202210. Morabito, P., (2001) Thermal Properties of Concrete Variations with the Temperature During the Hydration Phase Lulea University of Technology, SE 97187, Lulea Sweden, 23pp Morabito, P., (1998) Methods to Determin e Heat of Hydration o f Concrete, Prevention of Thermal Cracking in Concrete at Early Ages, RILEM Report 15, Edited by R. Springenschmid, E&FN Spon, London Neville, A., (2003) Neville on Concrete American Concrete Institute, Farmington Hills, Michigan Neville, A., (1995) Properties of Concrete 4th ed., Essex, England: Pearson Education Limited. Ng, P.L., Ng, I.Y.T., & Kwan, A.K.H. (2008) Heat Loss Compensation in Semi -Adiabatic Curing Test of Concrete ACI Materials Journal, 105(1) 5261 NT Build 480, (1997) Cement: Heat of Hydration, NORDTEST, Espoo, Finland, 6pp Ochiai, E., (2009) Thermal testing of Mass Concrete, Report University of Florida, To be published Odler, I. (1998) Hydration, Leas Chemistry of Cement and Concrete, Edited by Hewlett P.C., Burlington, Massachuse tts: Elsevier, 241 297 Ott, L., (1988) An Introduction to Statistical Methods and Data Methods, PWS -Kent, Boston

PAGE 273

273 Patel, H.H., Bland, C.H., & Poole, A.B. (1994) The Microstructure of Concrete Cured at Elevated Temperatures, Cement and Concrete Research 25(3) 485490 Peng, Y., Borgnakke, C., Hansen, W., & Biernacki, J. (2009) Mechanism and Modeling of Cement Hydration, Presented at the International Summit of Hydration Kinetics, Quebec, CA, July 27 29 Philleo, R.E., (1966) Strength Elastic Properties and C reep, ASTM STP 169, pp. 160175 Pirtz, D., & Carlson, R.W., (1963) Tests of Strain Meters and Stress Meters under Simulated Field Conditions, Symposium on Mass Concrete ACI SP 6, pp. 287308 Poole, J.L., Riding, K.A., Folliard, K.J., Juenger, M.C.G., & Sc hindler, A.K. (2007) Hydration Study of Cementitious Materials Using Semi -Adiabatic Calorimetry, Edited by Wang, K. & Schindler, A.K. Concrete Heat Development: Monitoring, Prediction, and Management ACI SP 241, 59 76 Poole, J.L., Riding, K.A., Folliar d, K.J., Juenger, M.C.G., & Schindler, A.K. (2007a) Methods for Calculating Activation Energy for Portland Cement ACI Materials Journal 104(11) 303311 Poole, J.L., (2007) Modeling Temperature Sensitivity and Heat Evolution of Concrete, PhD Dissertation, University of Texas, Austin Poole, T. (2007) Revision of Test Method and Specificiations for Controlling Heat of Hydration in Hydraulic Cement R&D Serial No. 3007, Portland Cement Association, Skokie, Illinois, 40 pp Poole, T. (2004) Predicting heat of h ydration in hydraulic cement, US Army Research and Development Center, Vicksburg, MS, Apr Poppe, A. & DeSchutter, G. (2005) Cement Hydration in the Presence of High Filler Contents, Cement and Concrete Research, 25 22902299 Price, W.H. (1974) Practical Qualities of Cement Journal of the American Concrete Institute 7 1, No. 9, September, pp 436444 Principles of Thermal Analysis and Calorimetry (2002), Edited by Haines, P.J., Cambridge, UK: The Royal Society of Chemistry Ramlochan, T., Zacarias, p., Thomas, M.D.A., and Hooton, R.D., (2003) The Effect of Pozzolans and Slag on the Expansion of Mortars Cured at Elevated Temperature Part I: Expansive Behaviour Cement & Concrete Research 33, 807814 Ramlochan T.,(2003) The Effect of Poz zolans and Slag on the Expansion of Mortars Cured at Elevated Temperatures, P hD Thesis, University of Toronto

PAGE 274

274 Reinhardt, H.W., Rossi, P, & VanMier, J.G.M. (1990) Joint Investigation of Concrete at High Rates of Loading, Materials and Structures 23, 213216 Riding, K.A., Poole, J.L., Schindler, A.K., Juenger, M.C.G., & Folliard, K.J. (2007) Temperature Boundary Condition Models for Concrete Bridge Members. ACI Materials Journal 104(4) 379387 Riding, K.A., Poole, J.L., Juenger, M.C.G., Schindler, A.K, & Fo lliard, K.J. (2007) Calorimetry Performed on Site: Methods and Uses. ACI SP 2413 Concrete Heat Development: Monitoring, Prediction and Management. Technical Session, April 22 26, 2007, Atlanta, Georgia, 25 38 Riding, K.A., Poole, J.L., Schindler, A.K., Ju enger, M.C.G., & Folliard, K.J. (2006) Evaluation of Temperature Prediction Methods for Mass Concrete Members, ACI Materials Journal 103(5), 357365 RILEM 119 TCE 1 (1997) Adiabatic and Semi -Adiabatic Calorimetry to Determine the Temperature Increase in Concrete due to Hydration Heat of Cement, Materials and Structures 30, 451457 Robbins, M., Predicting Temperature Rise and Thermal Cracking in Concrete, SN3030, Portland Cement Association, Skokie, Illinois, 2007, 194pp Ross, C.A., Jerome, D.M., Tedesco, J.W., & Hughes, M.L., (1996) Moisture and Stain Rate Effects on Concrete Strength ACI Materials Journal 9 3( 3), 293300 Roy, D.M.,White, E.L., & Nakagawa, Z. (1985) Effects of Early Heat of Hydration and Exposure to Elevated Temperatures on Properties of Mortars and Pastes with Slag Cement. Temperature Effects on Concrete Edited by Naik, ASTM STP 858, pp. 150167 Sargious, M., (1978) Performance of Transverse Joints in Prestressed and Reinforced Concrete Airfield Pavements Journal of the American Concrete Institute 75(8) 359366 Schindler, A.K. (2002) Concrete Hydration, Temperature Development, and Setting at Early Ages. PhD Dissertation, University of Texas at Austin, Austin, TX Schindler, A.K. (2004) Effect of Temperature on Hydration of Cementitious Materials. ACI Materials Journal, 101(1) 7281 Schin dler, A.K., & Folliard, K.J. (2005) Heat of Hydration Models for Cementitious Materials ACI Materials Journal, 102(1) 2433 Schlorholtz, S., and .Bergeson, K.L., (1993) Evaluation of Chemical Durability of Iowa Flyash Concrete, Iowa Department of Tra nsport ation, Ames, IA, March, Shi, C., & Day, R.L. (1993) Acceleration of Strength Gain of Lime-Pozzolan Cements by Thermal Activation. Cement and Concrete Research, 23 824 832.

PAGE 275

275 Skalny, J., Marchand, J., and Older, I., (2002) Sulfate Attack on Concrete, Spon P ress, London Stutzman, P. (2007) Relative Bias in Bogue C alculated Cement Compositions, Symposium on Test Methods for Chemical and Mineralogical Analysis of Hydraulic cements and Other Concrete -Making Materials Tampa, Florida Tank, R.C. (1988) The Rate Constant Model for Strength Development of Concrete, PhD Dissertation, Polytechnic University, NY Taylor, H.F.W., (1997) Cement Chemistry Heron Quay, London: Thomas Telford Publishing Thomas, M., Folliard, K., Drimala, T., & Ramlochan, T. (2008) Diagnosing Delayed Ettringite Formation in Concrete Structures. Cement and Concrete Research, 38 841-847 Torres, S.M., Lynsdale, C.J., Swamy, R.N., & Sharp, J.H. (2006) Microstructure of 5 Year Old Mortars Containing Limestone Filler Damaged by Thaumasitem Cemen t and Concrete Research, 36 384394 Tsivilis, S., Tsantilas, J., Kakali, G., Chaniotakis, E., & Sakellariou, A., (2001) The Permeability of Portland Limestone Cement Concrete, Cement and Concrete Research, Vol. 33 14651471 Tyner, M., (1946) The Effect of Moisture on the Thermal Conductivity of Limerock Concrete Journal of the American Concrete Institute, 18(1) 918 Ugur, I., & Demirdag, S. (2006) Investigating of the Relation between the Specific Heat Capacity and Material Properties of Some Natural B uilding and Facing Stones International Journal of Rock Mechanics and Mining Sciences, Vol. 43 831835 Verbeck, G.J., & Helmuth, R.H., (1986) Structure and physical properties of cement paste, in Proceedings of the Fifth International Symposium on the Chemistry of Cement, Part III, Tokyo, 1 32 Voigt, T., Sun, Z., & Shah, S.P. (2006) Comparison of Ultrasonic Wave Reflection Method and Maturity Method in Evaluating Early -Age Compressive Strength of Mortar. Ce ment & Concrete Composites, 28, 307316 Volz, C.K., Tucker, R.L., Burns, N.H., & Lew, H.S. (1980) Maturity Effects on Concrete Strength. Cement and Concrete Research, 11 41 50 Wadso, L (2003) An Experimental Comparison Between Isothermal Calorimetry, Semi Adiabatic Calorimetry, and Solution Calorimet ry for the Study of Cement Hydration, Nordtest Technical Report TR 522, 44pp Wadso I, Goldberg R.N., (2001) Standards in isothermal microcalorimetry (IUPAC technical report) Pure Appl Chem 73: 162539

PAGE 276

276 Wadso, L. (2001) Temperature Changes within Samples i n Heat Conduction Calorimeters, Thermochimica Acta, Vol 366 121127 Woods, H., Steinour, H.H. and Starke, H.R. (1933) Heat ev olved by cement in relation to S trength, Engineering News Record, Vol 110, p 431. Xu, A., Sarkar, S.L, & Nilsson, L.O. (1993) Eff ect of Flyash on the Microstructure of Cement Mortar. Materials and Structures 26, 414424 Zheng, X, ( 2007) Characterization of Fracture Properties of Concrete Using the Superpave IDT Test PhD Dissertation, University of Florida http://archives.math.utk.edu/ICTCM/VOL13/C013/paper.html last accessed December 7 2009

PAGE 277

277 BIOGRAPHICAL SKETCH Christopher C. Ferraro was born March 1, 1974, in Huntington, New York, to Ronald and Victoria Ferraro. He graduated from Walt Whitman High School, in Suffolk County, N.Y.in June of 1992. He received his Associate of Arts degree in December of 1994 from Palm Beach Community College, and transferred to the University of Florida to pursue a Bachelor of Science in Civil Engineering in the summer of 1995. While attending the University of Florida full time, Christopher worked part time for the Department of Civil Engineering, for two years as a teaching assistant and research assis tant for Dr. Fazil T. Najafi. He received his Bachelor of Science in Civil Engineering in May of 1998, graduating with honors. Upon graduation, he relocated to Long Island, New York, and worked full time in Manhattan, New York, as an Engineer Intern for Law Engineering Inc. After a year with Law, he transferred to STV Incorporated where he worked on several inspection projects under the supervision of Ms. Marjorie M. Lynch, P.E. who aided him in gaining expertise and knowledge which inspired him to resume his education. Christopher continued with his education, entering graduate school to pursue a Master of Engineering in the Materials Group of the Civil and Coastal Engineering Department in August 2000 under the advisement of Dr. Andrew Boyd. After gradua ting from the University of Florida with a Master of Engineering in 2003, Christopher continued his education and began work on a PhD with the University of Florida, while simultaneously working full time at the Florida Department of Transportation State M aterials Office (FDOT SMO) under the advisement of Dr. Mang Tia Upon the completion of this PhD Christopher intends to work as a post doctorate faculty member in the Department of Civil and Coastal Engineering at the University of Florida.