EVALUATION OF THE MATURITY METHOD TO ESTIMATE CONCRETE STRENGTH Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. ________________________________________ Samuel Allen Wade Certificate of Approval: _________________________ ________________________ Robert W. Barnes Anton K. Schindler, Chair Associate Professor Gottlieb Assistant Professor Civil Engineering Civil Engineering _________________________ ________________________ Mary L. Hughes Stephen L. McFarland Assistant Professor Acting Dean Civil Engineering Graduate School EVALUATION OF THE MATURITY METHOD TO ESTIMATE CONCRETE STRENGTH Samuel Allen Wade A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama December 16, 2005 iii EVALUATION OF THE MATURITY METHOD TO ESTIMATE CONCRETE STRENGTH Samuel Allen Wade Permission is granted to Auburn University to make copies of this thesis at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. ____________________________ Signature of Author ____________________________ Date of Graduation iv VITA Samuel Allen Wade, son of Don Edward and Carol Linda (Cronier) Wade, was born February 27, 1981, in Mission Viejo, California. He graduated from Newark High School, in Newark, Delaware, in 1999. He entered Auburn University in September of 1999 and received a Bachelor of Civil Engineering, magna cum laude, in December, 2003. He began Graduate School in January, 2004. v THESIS ABSTRACT EVALUATION OF THE MATURITY METHOD TO ESTIMATE CONCRETE STRENGTH Samuel Allen Wade Master of Science, December 16, 2005 (B.C.E., Auburn University, 2003) 361 Typed Pages Directed by Anton K. Schindler The strength of a newly constructed concrete structure or roadway is critically important to contractors and engineers who must decide when to safely allow construction loads, prestressing operations, or opening to traffic loads. The maturity method allows the user to estimate concrete strength at any given time using the structure?s unique temperature history. The purpose of this project was to analyze the effectiveness and accuracy of the maturity method to estimate concrete strengths for a variety of commonly used mixtures. The mixtures were chosen to show the effects of using various types of cements, various types and doses of supplementary cementing materials, and various water-to-cementitious materials ratios. Each mixture was batched at three different temperatures selected to span the entire range of expected conditions; vi with average batch temperatures of 55?F, 70?F, and 101?F. The 70?F batch was cured isothermally at normal laboratory temperatures (68?F-73?F). The 55?F and 101?F batches were cured in water baths that cycled between 40?F - 55?F and 90?F - 106?F, respectively. From each batch, nineteen 6x12 in. cylinders were prepared. Eighteen were used for compressive strength testing at 6 different ages. One cylinder was used to record the temperature history of the batch. Compressive strength versus age data was examined. Mixtures were evaluated based on the amount of long-term strength loss or gain due to curing temperatures and the amount of ?temperature sensitivity.? The time- temperature histories were then converted to maturity using the Nurse-Saul and Arrhenius maturity functions. Compressive strength versus maturity data were then analyzed to determine the accuracy of estimating strengths using the maturity method. It was found that the maturity method was inaccurate for estimating concrete strengths beyond 7 days of equivalent age, especially for mixtures with severe long-term strength loss due to high curing temperatures. A ?Modified ASTM? method was proposed to handle the late-age strength problems. Then, the mixtures were evaluated using a simplified approach employing constant values of temperature sensitivity. Finally, temperature sensitivity functions were proposed based on the concrete curing temperature. The results of this study indicated that the Nurse-Saul maturity function using a datum temperature of 32?F (0?C) was the most effective and practical method for estimating concrete strengths of all mixtures studied. However, if the Arrhenius function will be used, a temperature sensitivity function is recommended to allow for a higher temperature sensitivity at low curing temperatures and a lower temperature sensitivity at high curing temperatures. vii ACKNOWLEDGMENTS The author would like to thank Jeffrey Nixon for his assistance with all laboratory testing and preparations and Dr. Anton K. Schindler for his technical guidance. The author is grateful for the project funding provided by the Alabama Department of Transportation. Billy Wilson, Curtis Williams, and all undergraduate lab assistants also deserve thanks, for without them this project would never have been completed. Special thanks are due to the author?s mother and father; sister, Wendy Clare (Wade) McGuire; brother, Sergeant Edward Thomas Wade, United States Army, retired; and Michelle Elizabeth Foley, all of whom provided special encouragement throughout this project. viii Style Manual used: The Chicago Manual of Style, 15th edition Computer Software used: Microsoft Word, Microsoft Excel ix TABLE OF CONTENTS LIST OF FIGURES??????????????????????????.. xiii LIST OF TABLES??????????????????????????? xix CHAPTER 1: INTRODUCTION 1.1 Background???????????????????????????.. 1 1.2 Research Significance???????????????????????.. 2 1.3 Project Objectives????????????????????????? 3 1.4 Report Scope??????????????????????????? 4 CHAPTER 2: LITERATURE REVIEW 2.1 Concept and Definition of Maturity??????????????????.5 2.2 Use of the Maturity Method to Estimate Concrete Strength????????..11 2.2.1 Strength-Maturity Relationships????????????????... 11 2.2.2 Developing a Mixture-Specific Strength-Maturity Relationship????..14 2.2.3 Estimating Concrete Strength Using the Strength-Maturity??????. 15 Relationship 2.2.4 Examples from the Industry??????????????????.. 16 2.3 Limitations of the Maturity Method?????????????????. 22 2.3.1 Mixture-Specific Strength-Maturity Curve????????????...22 2.3.2 Effects of Curing Temperature on Long-Term Strength???????.. 25 x 2.3.3 Other Factors Affecting Concrete Strength????????????.. 31 2.4 Comparison of Maturity Functions?????????????????? 34 2.4.1 Maturity Function Trends Compared to Concrete Behavior?????? 35 2.4.2 Temperature Sensitivity Values????????????????? 45 2.5 Summary???????????????????????????? 51 CHAPTER 3: LABORATORY TESTING PROGRAM 3.1 Concrete Mixtures Considered???????????????????.. 53 3.2 Research Approach???????????????????????? 55 3.3 Raw Materials?????????????????????????? 58 3.3.1 Cements and Supplementary Cementing Materials?????????.. 58 3.3.2 Aggregates????????????????????????? 62 3.3.3 Chemical Admixtures????????????????????... 65 3.4 Mixture Proportions???????????????????????... 66 3.5 Summary???????????????????????????? 69 CHAPTER 4: LABORATORY PROCEDURES AND TESTS 4.1 Concrete Production Procedures??????????????????? 70 4.1.1 Mixing Procedures?????????????????????? 70 4.1.2 Mixing with Silica Fume???????????????????.. 73 4.2 Fresh Concrete Property Testing??????????????????... 73 4.3 Hardened Concrete Testing ????????????????????...74 4.4 Summary???????????????????????????? 77 xi CHAPTER 5: PRESENTATION OF RESULTS 5.1 Fresh Concrete Property Results???????????????????78 5.2 Hardened Concrete Test Results??????????????????? 82 5.2.1 Crossover Effect???????????????????????95 5.2.2 Temperature Sensitivity???????????????????? 97 5.3 Summary???????????????????????????.. 100 CHAPTER 6: ANALYSIS OF RESULTS 6.1 Analysis of Data Based on the ASTM C 1074 Method?????????...105 6.1.1 Nurse-Saul Maturity Function????????????????? 105 6.1.2 Arrhenius Maturity Function?????????????????...115 6.1.3 Accuracy of Maturity Functions Based on the ASTM C 1074????...121 Method 6.2 Modifications to the ASTM C 1074 Method?????????????... 137 6.2.1 Nurse-Saul Maturity Function????????????????? 141 6.2.2 Arrhenius Maturity Function?????????????????...144 6.2.3 Accuracy of Maturity Functions Based on the Modified ASTM???... 148 Method 6.3 Simplified Maturity Approach???????????????????. 164 6.3.1 Nurse-Saul Maturity Function????????????????? 169 6.3.2 Arrhenius Maturity Function?????????????????...177 6.4 Variable Temperature Sensitivity Models??????????????... 185 6.5 Summary and Conclusions????????????????????...189 xii CHAPTER 7: CONCLUSIONS AND RECOMMENDATIONS 7.1 Conclusions??????????????????????????.. 194 7.2 Recommendations???????????????????????? 198 7.3 Summary???????????????????????????.. 199 REFERENCES???????????????????????????? 200 APPENDICES????????????????????????????. 204 APPENDIX A: Compressive Strength Test Results?????????????...205 APPENDIX B: Strength-Maturity Relationships for Nurse-Saul Maturity????...212 Function, ASTM C 1074 Method APPENDIX C: Strength-Maturity Relationships for Arrhenius Maturity?????.223 Function, ASTM C 1074 Method APPENDIX D: Error Tables for ASTM C 1074 Method???????????...229 APPENDIX E: Strength-Maturity Relationships for Nurse-Saul Maturity????... 249 Function, Modified ASTM Method APPENDIX F: Strength-Maturity Relationships for Arrhenius Maturity?????. 255 Function, Modified ASTM Method APPENDIX G: Error Tables for Modified ASTM Method???????????261 APPENDIX H: Error Tables for Nurse-Saul Maturity Function, Simplified????. 301 Method APPENDIX I: Error Tables for Arrhenius Maturity Function, Simplified????... 321 Method xiii LIST OF FIGURES Figure 2.1: Results from Nurse?s experiments (1949)????????????... 6 Figure 2.2: Diagram of concrete maturity using Nurse-Saul maturity??????... 8 function Figure 2.3: Diagram of Saul?s maturity rule????????????????.. 9 Figure 2.4: Converting strength-age data to strength-maturity data???...???..11 Figure 2.5: Various strength-maturity functions (Carino 1991)????????...14 Figure 2.6: Example of estimating strength from strength-maturity??????? 19 relationship (TxDOT 1999) Figure 2.7: Example of verification of design strength (TxDOT 1999)?????...20 Figure 2.8: Example of verification of safety or formwork strength??????... 21 Figure 2.9: Further results from Nurse?s experiments (1949)?????????.. 23 Figure 2.10: Ultimate compressive strength (Su) values for Carino and ?????... 25 Lew?s experiment (1983) Figure 2.11: Temperature affects on long-term strength (Saul 1951)??????? 27 Figure 2.12: Temperature effects on estimating strength using?????????. 28 strength-maturity relationship (Saul 1951) Figure 2.13: Cross-over effect due to curing temperatures (Alexander and????.. 29 Taplin 1962) Figure 2.14: Effect of air content on concrete strength (Cordon 1946)??????. 32 Figure 2.15: Effect of moist-curing time on strength gain of concrete??????.. 34 (Gonnerman and Shuman 1928) xiv Figure 2.16: Age conversion factors for Nurse-Saul and Arrhenius maturity???... 38 functions Figure 2.17: Rate constants used to find best-fit datum temperature???????. 40 (Carino 1991) Figure 2.18: Rate constants used to find best-fit activation energy???????... 41 (Carino 1991) Figure 2.19: Age conversion factors with best-fit Nurse-Saul and???????? 43 Arrhenius age conversion factors (Carino 1991) Figure 2.20: Two datum temperature scheme for Nurse-Saul maturity??????. 44 function (Carino 1991) Figure 2.21: Comparing age conversion factors using FHP model to other????.. 49 functions and values Figure 3.1: Curing tank and programmable heating/cooling circulator?????... 56 Figure 3.2: Inside of a temperature-controlled curing tank??????????.. 57 Figure 3.3: Curing tank cycles?????????????????????..58 Figure 3.4: Gradation test results for Martin Marietta Fine Aggregate??..???. 63 Figure 3.5: Gradation test results for Superior Products Fine Aggregate.....???... 64 Figure 3.6: Gradation test results for Martin Marietta No. 67 Limestone?..???. 64 Figure 3.7: Gradation test results for Vulcan Materials No. 78 Limestone????. 65 Figure 5.1: Typical temperature history of concrete cylinders?????................ 82 Figure 5.2: Compressive strength versus age results for Type I - 0.41?????? 84 mixture Figure 5.3: Compressive strength versus age results for Type I - 0.44?????? 85 mixture Figure 5.4: Compressive strength versus age results for Type I - 0.48?????? 85 mixture Figure 5.5: Compressive strength versus age results for 20% F mixture?????. 86 xv Figure 5.6: Compressive strength versus age results for 30% F mixture?????. 86 Figure 5.7: Compressive strength versus age results for 20% C mixture????? 87 Figure 5.8: Compressive strength versus age results for 30% C mixture????? 87 Figure 5.9: Compressive strength versus age results for 30% Slag mixture???? 88 Figure 5.10: Compressive strength versus age results for 50% Slag mixture???? 88 Figure 5.11: Compressive strength ve rsus age results for Type III - 0.37?????. 89 mixture Figure 5.12: Compressive strength versus age results for Type III - 0.44?????. 89 mixture Figure 5.13: Compressive strength versus age results for 70/20/10 - 0.37????? 90 mixture Figure 5.14: Compressive strength versus age results for 70/20/10 - 0.44?????90 mixture Figure 5.15: Various degrees of temperature sensitivity???????????? 98 Figure 6.1: Method used to find the best-fit datum temperature,???????... 107 ASTM C 1074 method for Type I ? 0.41 mixture Figure 6.2: 20% F mixture, (a) compressive strength vs. age,????????... 111 (b) strength-maturity plot, ASTM C 1074 method Figure 6.3: Type I - 0.41 mixture, (a) compressive strength vs. age,??????. 112 (b) strength-maturity plot, ASTM C 1074 method Figure 6.4: Type III - 0.37 mixture, (a) compressive strength vs. age,?????.. 113 (b) strength-maturity plot, ASTM C 1074 method Figure 6.5: Method for computing activation energy based on????????..117 ASTM C 1074 for the Type I - 0.41 mixture Figure 6.6: Strength-maturity plot, ASTM C 1074 method for 20% F?????.. 118 mixture Figure 6.7: Strength-maturity plot, ASTM C 1074 method for????????. 119 Type I - 0.41 mixture xvi Figure 6.8: Strength-maturity plot, ASTM C 1074 method for????????. 119 Type III - 0.37 mixture Figure 6.9: Estimated strength versus actual strengths for 20% F mixture???? 125 Figure 6.10: Estimated strength versus actual strengths for the????????... 126 Type I - 0.41 mixture Figure 6.11: Estimated strength versus actual strengths for the????????... 127 Type III - 0.37 mixture Figure 6.12: Estimated strength versus actual strengths for the 20% F?????... 128 mixture Figure 6.13: Estimated strength versus actual strengths for the????????... 129 Type I - 0.41 mixture Figure 6.14: Estimated strength versus actual strengths for the????????... 130 Type III - 0.37mixture Figure 6.15: Change in rate values and datum temperature for????????... 138 Type I - 0.41 mixture Figure 6.16: Strength-maturity plot, Modified ASTM method for 20% F????...143 mixture Figure 6.17: Strength-maturity plot, Modified ASTM method for???????.. 143 Type I - 0.41 Figure 6.18: Strength-maturity plot, Modified ASTM method for???????.. 144 Type III - 0.37 mixture Figure 6.19: Strength-maturity plot, Modified ASTM method for???????. 146 20% F mixture Figure 6.20: Strength-maturity plot, Modified ASTM method for???????.. 147 Type I - 0.41 mixture Figure 6.21: Strength-maturity plot, Modified ASTM method for???????.. 147 Type III - 0.37 mixture Figure 6.22: Error plot for the 20% F mixture, Modified ASTM???????? 150 method, NSM xvii Figure 6.23: Error plot for the 20% F mixture, ASTM C 1074 method,?????..150 NSM Figure 6.24: Error plot for the Type I - 0.41 mixture, Modified ASTM?????.. 152 method, NSM Figure 6.25: Error plot for the Type I - 0.41 mixture, ASTM C 1074 method,??... 152 NSM Figure 6.26: Error plot for the Type III - 0.37 mixture, Modified ASTM????... 154 method, NSM Figure 6.27: Error plot for the Type III - 0.37 mixture, ASTM C 1074?????...154 method, NSM Figure 6.28: Error plot for the 20% F mixture, Modified ASTM method,????...156 AM Figure 6.29: Error plot for the 20% F mixture, ASTM C 1074 method, AM???.. 156 Figure 6.30: Error plot for the Type I - 0.41 mixture, Modified ASTM?????.. 158 method, AM Figure 6.31: Error plot for the Type I - 0.41 mixture, ASTM C 1074??????.158 method, AM Figure 6.32: Error plot for the Type III - 0.37 mixture, Modified ASTM????... 160 method, AM Figure 6.33: Error plot for the Type III - 0.37 mixture, ASTM C 1074?????.. 160 method, AM Figure 6.34: Obtaining age conversion factors for Type I - 0.41 mixture?????166 Figure 6.35: Age conversion factors for all mixtures????????????... 168 Figure 6.36: Error plot for the 20% F mixture, simplified method,???????. 172 To = 14?F Figure 6.37: Error plot for the 20% F mixture, simplified method,???????. 172 To = 32?F Figure 6.38: Error plot for the Type I - 0.41 mixture, simplified method,????.. 174 To = 14?F xviii Figure 6.39: Error plot for the Type I - 0.41 mixture, simplified method,????... 174 To = 32?F Figure 6.40: Error plot for the Type III - 0.37 mixture, simplified method,???? 176 To = 14?F Figure 6.41: Error plot for the Type III - 0.37 mixture, simplified method,???? 176 To = 32?F Figure 6.42: Error plot for the 20% F mixture, simplified method,???????. 180 E= 40 kJ/mol Figure 6.43: Error plot for the 20% F mixture mixture, simplified method,????180 E = 25 kJ/mol Figure 6.44: Error plot for the Type I - 0.41 mixture, simplified method,????... 182 E= 40 kJ/mol Figure 6.45: Error plot for the Type I - 0.41 mixture, simplified method,????... 182 E = 25 kJ/mol Figure 6.46: Error plot for the Type III - 0.37 mixture, simplified method,???? 184 E= 40 kJ/mol Figure 6.47: Error plot for the Type III - 0.37 mixture, simplified method,???? 184 E = 25 kJ/mol Figure 6.48: Proposed temperature sensitivity functions???????????. 187 xix LIST OF TABLES Table 2.1: Datum temperature values proposed by Carino and Tank ?????? 46 (1992) Table 2.2: Activation energy values proposed by various research efforts???? 50 (Carino 1991) Table 2.3: Activation energy values proposed by Carino and Tank??????... 51 (1992) Table 3.1: Types of mixtures evaluated?????????????????.. 54 Table 3.2: Composition of cementitious materials?????????????. 61 Table 3.3: Specific surface areas and specific gravities of all????????? 62 cementitious material properties Table 3.4: Specific gravities and absorptions for all materials????????... 65 Table 3.5 (a): Mixture proportions for all concrete mixtures used????????.. 67 in this study Table 3.5 (b): Mixture proportions for all concrete mixtures used????????.. 68 in this study Table 4.1: Target slump and air content for all mixtures (ALDOT 2002)????. 71 Table 4.2 Compressive strength testing schedule?????????????.. 76 Table 5.1: Fresh concrete properties of all mixtures ????.???.????? 81 Table 5.2: Regression values for strength-age curves????????????.91 Table 5.3: Crossover factors and crossover time for all mixtures???????.. 94 Table 6.1: Best-fit datum temperatures based on ASTM C 1074???????. 108 xx Table 6.2: Datum temperature values proposed by Carino and Tank?????... 115 (1992) Table 6.3: Best-fit activation energies based on ASTM C 1074 methods???? 118 Table 6.4: Activation energy values proposed by Carino and Tank?????? 120 (1992) Table 6.5: Error using NSM function based on ASTM C 1074 methods????. 125 for the 20% F mixture Table 6.6: Error using NSM function based on ASTM C 1074 methods????. 126 for the Type I - 0.41 mixture Table 6.7: Error using NSM function based on ASTM C 1074 methods????. 127 for the Type III - 0.37 mixture Table 6.8: Error using AM function based on ASTM C 1074 methods????... 128 for the 20% F mixture Table 6.9: Error using AM function based on ASTM C 1074 methods????... 129 for the Type I - 0.41 mixture Table 6.10: Error using AM function based on ASTM C 1074 methods????... 130 for the Type III - 0.37 mixture Table 6.11: Comparison of maturity functions for all mixtures, ASTM????? 132 C1074 method Table 6.12: Regression values for strength-age curves up to 7 days of ?...???.. 140 equivalent age Table 6.13: Datum temperatures obtained from different methods??????? 141 Table 6.14: Activation energies based obtained from different methods????... 145 Table 6.15: Error using NSM function based on Modified ASTM method???... 149 for 20% F mixture Table 6.16: Error using NSM function based on Modified ASTM method???... 151 for Type I - 0.41 mixture Table 6.17: Error using NSM function based on Modified ASTM method???... 153 for Type III - 0. 37 mixture xxi Table 6.18: Error using AM function based on Modified ASTM method????. 155 for 20% F mixture Table 6.19: Error using AM function based on Modified ASTM method????..157 for Type I - 0.41 mixture Table 6.20: Error using AM function based on Modified ASTM method????..159 for Type III - 0.37 mixture Table 6.21: Comparison of maturity methods for all mixtures, Modified????.. 164 ASTM method Table 6.22: R2 values and rate constants at the reference temperature?????... 167 for all mixtures Table 6.23: Error using NSM function based on simplified method for????? 171 20% F mixture Table 6.24: Error using NSM function based on simplified method fo r????? 173 Type I - 0.41 mixture Table 6.25: Error using NSM function based on simplified method for????? 175 Type III - 0.37 mixture Table 6.26: Comparison of NSM functions for all mixtures, simplified????? 177 method Table 6.27: Error using AM function based on simplified method for?????...179 20% F mixture Table 6.28: Error using AM function based on simplified method for?????...181 Type I - 0.41 mixture Table 6.29: Error using AM function based on simplified method for?????.. 183 Type III - 0.37 mixture Table 6.30: Comparison of AM functions for all mixtures, simplified?????.. 185 method 1 CHAPTER 1 INTRODUCTION 1.1 BACKGROUND It has long been known that freshly placed concrete behaves differently in cold and hot environments. Concrete cured in high temperatures gains strength faster than concrete cured in low temperatures. This temperature dependency causes difficulty for contractors and engineers attempting to determine the strength of concrete placed and cured at various ambient conditions. Contractors need to know the strength of a structure or roadway in order to meet deadlines for formwork removal, traffic openings, transfer of prestress force, and other construction operations dependant on strength. The maturity method, developed progressively by McIntosh (1949), Nurse (1949), and Saul (1951), gives the contractor the ability to estimate concrete strength based on the time- temperature history of the concrete in question. This is very useful because current Alabama Department of Transportation (ALDOT) practices rely on the strength of laboratory-cured cylinders to ensure desired strengths have been achieved in the actual structure. This can cause unnecessary delays, since a structure?s strength gain may be significantly different than that of a test specimen due to differences in concrete volume and environmental effects. The maturity method allows the contractor to estimate the strength of a structure based on the temperature history of the structure using a predetermined strength relationship, called a strength-maturity relationship. 2 There have been numerous research efforts verifying the use of the maturity method to estimate concrete strength (Bergstrom 1953; Tank and Carino 1991). However, the maturity method must be used with caution. Mixture proportions must remain fairly constant, or the predetermined strength-maturity relationship becomes invalid (ASTM C 1074 2004). Also, the structure and representative cylinders must have ample moisture supplied for proper hydration of the concrete (ASTM C 1074 2004). Severe curing temperatures at early ages can also cause errors in strength estimations using the maturity method (Alexander and Taplin 1962; Carino 1991). This study evaluated the accuracy of the maturity method used to estimate concrete strength. Thirteen concrete mixtures were evaluated at three different curing temperatures. The mixtures were chosen to show the effect, with respect to strength gain, of various types and doses of supplementary cementing materials (SCMs), as well as varying cement types, and water-to-cementitious materials ratios. 1.2 RESEARCH SIGNIFICANCE Currently there are few state departments of transportation that have implemented specifications for estimating concrete strength using the maturity method. Some techniques can be learned from these specifications; however, the maturity method should be tested using methods and standards already used by ALDOT. Also, many past studies (Tank and Carino 1991; Carino and Tank 1992) evaluated the accuracy of the maturity method using concrete cured under isothermal temperature conditions. In order to better simulate field conditions, concrete tested in this study was cured under 3 fluctuating curing temperatures, except for control batches which were cured isothermally at normal laboratory temperatures (68?F - 73?F). There are two functions that are recommended for computing the maturity of concrete: the Nurse-Saul maturity function and the Arrhenius maturity function (ASTM C 1074 2004). Carino (1991) concluded that the Arrhenius maturity function estimates concrete strength better than the Nurse-Saul maturity function. However, in practice, state DOT?s tend to use the less complicated Nurse-Saul maturity function (Texas DOT 1999; Iowa DOT 2000). The mixtures used in this study will be evaluated using both functions, and recommendations based on the accuracy of both functions will be made for use on ALDOT projects. 1.3 PROJECT OBJECTIVES The objectives of this study were as follows: 1) Determine the effect of various fluctuating curing temperatures on concrete strength behavior. 2) Determine the effect of various types and doses of SCMs, varying cement types, and water-to-cementitious materials ratios on the rate of strength gain at different temperatures. 3) Determine the accuracy of the maturity method in estimating concrete strength for numerous mixtures with varying types and doses of SCMs, varying cement types, and water-to-cementitious materials ratios using ASTM C 1074 maturity methods. 4 4) Develop modifications to the current ASTM C 1074 procedure to handle any deficiencies, if necessary, and determine the improvement in strength prediction accuracy. 5) Determine the maturity function to be used on ALDOT projects that minimizes strength estimation errors while maintaining ease of use. 1.4 REPORT SCOPE Chapter 2 gives the technical background of the maturity method as well as describes in detail the methods used to estimate concrete strength on a construction project. Chapter 3 describes the mixtures used in this study as well as reports all raw materials and mixture proportions used. The research approach is also given, describing the plan of action. Chapter 4 details the experimental testing plan used for this study, including mixing procedures and tests performed on fresh and hardened concrete. Results of the testing are presented in Chapter 5. Also the effects of hot and cold fluctuating temperatures and the various types and doses of SCMs, varying cement types, and water-to-cementitious materials ratio on strength behavior are discussed. In Chapter 6, the accuracy of the maturity method in estimating concrete strengths is examined. First, the data are analyzed based on current ASTM C 1074 procedures. Then, modifications are made in an effort to improve accuracy. Next a simplified approach to using the maturity method is applied and finally, new models fitted to the strength data obtained in this study are proposed. Chapter 7 summarizes the study, offer conclusions, and presents recommendations for the methods to be used for ALDOT projects. 5 CHAPTER 2 LITERATURE REVIEW This chapter provides the historical and technical background required to understand the use, effectiveness, accuracy, and limitations of the maturity method for estimating concrete strength. 2.1 CONCEPT AND DEFINITION OF THE MATURITY METHOD The idea that the rate of concrete strength gain is based on the curing time and temperature history was first noted by McIntosh (1949). He hypothesized that the ?rate of hardening at any moment is directly proportional to the amount by which the curing temperature exceeds the [datum] temperature.? He defined this hardening index as ?basic age.? The datum temperature was defined as the temperature below which concrete will not harden, which McIntosh chose to be 30?F (-1?C). However, he found that specimens cured at higher temperatures, up to 200?F (93?C), did not have the same compressive strength versus basic age trend as specimens cured at control temperatures of 60?F (16?C). Therefore, the basic-age assumption did not hold true; however, McIntosh?s hypothesis was not completely unfounded. Soon after, Nurse (1949) published his findings on the effects of steam curing on concrete. In his study, Nurse cured different concrete mixtures at temperatures ranging from 64?F to 212?F (18?C to 100?C) and tested the compressive strength at various 6 times. In order to compare the effects of time and temperature on the compressive strength of the different mixtures evaluated, Nurse expressed the strengths ?as a percentage of the strength after 3 days? storage at normal temperature [64?F].? Then he plotted the strength percentages versus the product of the temperature at which the concrete had cured and the curing time. When the strength percentages were plotted against the temperature-time ?products? they followed a distinct curve. As the temperature-time product increased, the percent of 3-day compressive strength increased as well. Some of Nurse?s results are shown in Figure 2.1, where the x-axis is the temperature-time product (?C?hr), the y-axis is the percentage of 3-day strength, and the data sets?gravel aggregate and clinker B?are based on different mixture designs. However, Nurse found that other mixtures did not fall on the same strength versus temperature-time curve. Figure 2.1: Results from experiments performed by Nurse (1949) Temperature x Time (?C?hr) Percentage of 3 day Compressive Strength (at 64?F [18 ?C]) 7 Later, in a follow-up to the findings of Nurse, Saul (1951) defined the ???maturity? of concrete? as its age multiplied by the average temperature above freezing which it has maintained.? With the combination of Nurse?s time-temperature product and Saul?s maturity definition, the first maturity function was born. The Nurse- Saul maturity (NSM) function, as it is commonly known, is defined in ASTM C 1074 as follows: Nurse-Saul Maturity (NSM) Function: tTTM t oc ???= ? 0 )( Equation 2.1 where, M = maturity index at age t, (?F?hour or ?C?hour), Tc = average concrete temperature during the time interval, Dt, (?F or ?C), Dt = a time interval (hr), and To = datum temperature (?F or ?C). The Nurse-Saul maturity function computes a maturity index, called the ?Temperature-Time Factor.? The maturity index is the quantitative amount of temperature and time a concrete mixture has accumulated. Computation of the Nurse- Saul maturity function can be explained by Figure 2.2. Basically, any time interval during which a concrete cures above the datum temperature cumulatively adds to the maturity index. Saul (1951) suggested using To = 32?F (0?C), but mentioned using a 8 lower datum temperature if concrete will be subjected to lower temperatures after setting. More discussion on choosing proper datum temperatures will be given in Section 2.4. Figure 2.2: Diagram of concrete maturity using Nurse-Saul maturity function Saul (1951) also defined the ?maturity rule,? stating that: Concrete of the same [mixture] at the same maturity (reckoned in temperature-time) has approximately the same strength whatever combination of temperature and time go to make up that maturity. This rule is the basis of using the ?maturity method? for estimating concrete strength. It means that, for a given concrete mixture, as long as the maturity index corresponding to a particular strength has been established, one can estimate when a concrete will reach that strength, even if the concrete of interest has a different curing history than the original concrete. A schematic of this concept is given in Figure 2.3. The figure shows the concept that, for the same mixture, concrete cured longer in a cold environment can have the same maturity index as concrete cured for a shorter period in a hot environment. Time (hr or d) Concrete Temperature ( ?F) Datum Temperature = To Maturity Index (Nurse-Saul Maturity Function) 9 According to Saul?s maturity rule, the strength of these hot and cold batches of concrete will be approximately the same at that maturity. Figure 2.3: Diagram of Saul?s maturity rule The Nurse-Saul maturity function is not the only function commonly used to compute a maturity index. Another function was introduced by Freiesleben Hansen and Pedersen (1977) and it is known as the Arrhenius maturity function. The maturity index that this function produces is called equivalent age. The Arrhenius maturity function, given in SI units, is defined in ASTM C 1074 as follows: Arrhenius Maturity (AM) Function: tet t TTRE e rc ??= ? ?? ? ?? ? +?+ ? 0 273 1 273 1 Equation 2.2 where, te = equivalent age at the reference curing temperature (hr), Tc = average concrete temperature for the time interval, Dt, (?C), M 1 = M 2 Cold Hot Time To Temperature Time M1 M2 Compressive Strength Temperature-Time Factor 10 Tr = reference temperature, (usually either 20?C or 23?C), E = activation energy, J/mol, R = universal gas constant, 8.314 J/(mol?K), and Dt = a time interval (hr). The reference temperature is rather arbitrary, and usually either 20?C or 23?C is used. It serves as a neutral point with respect to temperature effects on concrete strength gain behavior. The Arrhenius maturity function is based on the same idea as Saul?s maturity rule: once the time needed to reach a certain strength for a particular mixture is known at the reference curing temperature, the equivalent age to reach that strength will be the same, no matter the curing history. The Arrhenius maturity function accounts for time intervals that the concrete cures above or below the reference temperature, 73?F (23?C) in this study. If a mixture cures above the reference temperature, less actual time will be needed to reach the appropriate strength and if a mixture cures below the reference temperature, more actual time will be required to reach the same strength; however, the equivalent ages will be the same. The activation energy is an unknown in Equation 2.2, and it has to be selected to characterize the temperature sensitivity of the mixture. It has to do with the energy required to begin the hardening process for a particular mixture (Carino 1997). Using an appropriate activation energy has a great deal to do with the accuracy of strength estimations using the Arrhenius maturity function (ASTM C 1074 2004). This will be examined further, along with using proper datum temperatures, in Section 2.4. The next section will describe the process of establishing a strength versus maturity relationship 11 and how to estimate strengths using a maturity function, in other words, how to estimate strength using the maturity method. 2.2 USE OF MATURITY METHOD TO PREDICT CONCRETE STRENGTH 2.2.1 Strength-Maturity Relationships Figure 2.4 shows how maturity functions convert concrete of the same mixture, cured at different temperatures, to one compressive strength versus maturity index plot. First, the temperature history of the hardening concrete is measured at regular intervals. Through the use of one of the maturity functions defined earlier, the maturity can be determined. In Figure 2.4, the cold- and hot-cured mixtures? test strengths shifted toward the lab-cured strength-time curve once the temperature histories were converted to maturity. Note that the data points only shift on the time axis; that is, the strength levels are not altered. 0 2,000 4,000 6,000 8,000 0 400 800 1,200 Concrete Age (hours) Compressive Strength (psi) Lab Cure Cold Cure Hot Cure 0 10,000 20,000 30,000 40,000 Maturity Index (?F?hr) Lab Cure Cold Cure Hot Cure Figure 2.4: Converting strength-age data to strength-maturity data Strength-Maturity Relationship 12 The resulting strength versus maturity index plot is known as the strength- maturity relationship for a particular mixture (ASTM C 1074 2004). This relationship describes the strength of a particular concrete mixture at any maturity index. Accuracy of the strength-maturity relationship is based on the difference between the calculated strength from the relationship and the actual test strengths at the same maturity. There are many functions that have been proposed by different researchers to model these strength-maturity relationships, but three have received the most attention. According to Carino (1991), they are as follows: A) Exponential Function: a M u eSS ???????= t (Equation 2.3) where, S = compressive strength at maturity M, (psi), Su = limiting compressive strength (psi), M = maturity index (?F?hr or hr), t = characteristic time constant (?F?hr or hr), and a = shape parameter. B) Logarithmic Function: )log( MbaS += (Equation 2.4) where, a,b = constants (mixture dependent). 13 C) Hyperbolic Function: )(1 )( o o u MMk MMkSS ?+ ?= (Equation 2.5) where, Mo = maturity index when strength development is assumed to begin (?F?hr or hr), and k = rate constant, initial slope of strength-maturity curve (1/(?F?hr) or 1/hr). In order to determine which function best describes concrete strength gain, Carino fit the formulas to compressive strength data he obtained from a Type I mixture with a water-to- cement ratio of 0.45, cured at 73?F. His results are shown in Figure 2.5. The compressive strength results are reported in MPa, and the equivalent age has units of days. The hyperbolic and exponential equations produce almost identical curves, with slight variations in late-age strengths. These two equations (Equations 2.5 and 2.3) fit the data very well. The logarithmic equation (Equation 2.4) slightly underestimates strength around 5 days and ?predicts ever increasing strength with increasing maturity? at late- ages, which is one of the major criticisms of the function (Carino 1991). 14 Figure 2.5: Various strength-maturity functions (Carino 1991) 2.2.2 Developing a Mixture-Specific Strength-Maturity Relationship Now that the concept of a strength-maturity relationship has been introduced, its use with respect to actual construction practices can be explained. First, the strength- maturity relationship for the particular concrete mixture of interest must be established. This procedure is outlined in ASTM C 1074 (2004). First, a sample batch of concrete is prepared prior to construction of the actual roadway or structure. At least 15 cylinders are prepared. The cylinders are made according to ASTM C 192 (2004). Temperature sensors are placed in two of the cylinders to record the temperature history of the concrete. Acceptable temperature sensors include thermocouples, digital data-loggers, or commercial maturity meters. The sensors must be able to record temperature at least Test Data Equation 2.4 Equation 2.5 Equation 2.3 (Days) 15 once every 30 min for the first 48 hr and at least once every hour after that. Also, they must be able to accurately read temperatures to within ?1.8?F (1?C). The cylinders are moist-cured, and compressive strength tests are performed at 1, 3, 7, 14, and 28 days. The compressive strengths are tested according to ASTM C 39 (2004). If more cylinders were prepared from the sample batch, then additional testing times should fit estimated ages and strengths of interest for the actual structure. The strength and maturity at each test age is recorded. Maturity may be computed using either the Nurse-Saul or Arrhenius maturity functions. The data points are plotted on a strength versus maturity index graph. ASTM C 1074 (2004) allows a curve to be drawn through the data to establish the strength-maturity relationship; however, a regression analysis using any of the strength-maturity functions given in Equations 2.3 through 2.5 will give more accurate strength estimations. 2.2.3 Estimating Concrete Strength Using a Strength-Maturity Relationship Once the unique strength-maturity relationship for the concrete of interest has been determined, strength can be estimated on the job-site. According to ASTM C 1074 (2004), temperature sensors are inserted in the structure of interest after the concrete has been placed, but before initial setting occurs. Sensors should be placed in critical locations at which strength estimations are most desired. When the strength at a sensor location is desired, the maturity index (using either Nurse-Saul or Arrhenius maturity functions) of the concrete at that location should be calculated and recorded. Then the estimated strength can be obtained from the strength-maturity relationship that was previously determined. Before this estimated strength can be accepted, the estimated 16 strength must be validated by other methods. ASTM C 1074 recommends several methods of validation. These include in-place tests such as the ?Standard Test Method for Penetration Resistance of Hardened Concrete? (ASTM C 803), ?Standard Test Method for Compressive Strength of Concrete Cylinders Cast in Place in Cylindrical Molds? (ASTM C 873), or ?Standard Test Method for Pullout Strength of Hardened Concrete? (ASTM C 900); other methods such as ?Standard Test Method for Measuring Early-Age Compressive Strength and Projecting Later-Age Strength?(ASTM C 918) or ?Standard Test Method for Making, Accelerated Curing, and Testing Concrete Compression Test Specimens? (ASTM C 684); or field-prepared cylinders, made from the same concrete as was used in the structure, that are compared to the estimated strength found from the strength-maturity relationship. 2.2.4 Examples from the Industry Now that the standard procedure for estimating concrete strength using the maturity method has been presented, it may be helpful to present some specific techniques currently in use by other state departments of transportation (DOTs). First, the Texas Department of Transportation (TxDOT) specification (1999) for estimating concrete strength based on the maturity method will be discussed. For developing the strength-maturity relationship, TxDOT follows the methods outlined in ASTM C 1074, or Section 2.2.2 in this report, with the following changes or additions: ? Use of the Nurse-Saul maturity function with datum temperature of 14 ?F (-10?C), ? Calibration of temperature sensors before use on a project or at least annually, ? Use a minimum batch size of 4 yd3 to establish the strength-maturity relationship, 17 ? Use of logarithmic strength-maturity equation as defined in Equation 2.4, ? Minimum allowable coefficient of determination (R2) value of 0.9 for logarithmic equation, and ? Increased inspection of batching and raw materials. TxDOT?s procedure for estimating strength using the maturity method is the same as outlined in ASTM C 1074, or Section 2.2.3 in this report, except the temperature sensors are attached to, but not in direct contact with, reinforcing steel or formwork prior to concrete placement. TxDOT identifies critical temperature sensor locations as thin sections of slabs, concrete around steel tendons, concrete subjected to the worst environmental effects, and/or the last section of concrete placed in a day. TxDOT requires a new strength-maturity relationship to be determined if there are any changes in the mixture proportions including ?change in type, source, or proportion of cement, fly ash, coarse aggregate, fine aggregate, or admixtures?[or] a change in water-to- cementitious material ratio greater than 0.05? (TxDOT 1999). ASTM C 1074 also warns of changes in mixture proportions, but does not give specific guidance on when to establish a new relationship. For large projects, TxDOT recommends verifying the strength-maturity relationship a minimum of either once every 600 yd3 for structures and 30,000 yd3 for roadways, or once a month. For verification, TxDOT utilizes the last option provided by ASTM C 1074, which uses field-prepared and laboratory-cured cylinders to verify the strength-maturity relationship. TxDOT?s specification explains the verification procedure as follows: For verifying design strength, the specimens will be broken when the TTFs of the specimens are at least equal to the Required TTF of the member. For verifying strength for safety- or structurally-critical 18 formwork or falsework or steel stressing operations, the specimens will be broken when the Required TTF of the member is achieved, regardless of the TTFs of the specimens at that time. Recall that TTF refers to the maturity index obtained from the Nurse-Saul maturity function (Temperature-Time Factor). There is a distinct difference in these two types of verifications. The former statement means that the test cylinders must have the same maturity index required in the structure and therefore must have the required design strength. Thus, the cylinder may need considerably more time to reach the maturity index that the structure has achieved, due to the difference in mass of concrete and curing effects. The later statement means that the estimated safety or formwork strengths on the structure are sufficient as long as the cylinder?s test strengths agree with the strength- maturity relationship, at the cylinder?s maturity index. This is true even if the cylinder?s test strength and maturity are less than the structure?s strength and maturity. An example of TxDOT?s method of estimating concrete strength from the logarithmic strength-maturity function is shown in Figure 2.6. Again, they are using the Nurse-Saul maturity function and have determined that the structure or roadway needs to achieve a maturity index of 2,615 ?C?hr to reach the required strength of 3,600 psi. The process for verifying design strengths, as described above, is shown in Figure 2.7. The cylinder must have achieved a maturity index within the 10% tolerance, in this case between 2350 ?C?hr and 2880 ?C?hr, and the test strength must be within 10% of the value obtained from the strength-maturity curve, 3,600 psi for this example. Finally, an example of the verification method used for safety, formwork, and prestress transfer strengths is shown in Figure 2.8. This was not shown in TxDOT?s specification, but it has been included in this report in order to fully explain this method. The author has 19 attempted to keep the format the same as the TxDOT examples. In Figure 2.8, assume the strength required for safety was 4,200 psi, which corresponds to a maturity index of about 4,000 ?C?hr in the structure. When the structure reached this maturity (determined from its time-temperature history), suppose the cylinder maturity was only 2,000 ?C?hr. In order to verify the safety, formwork, or prestress transfer strength, the cylinder strength must be within ?10% of the strength indicated by the strength-maturity curve? or in this case approximately 3,400 psi ?340 psi. In this example, the tested cylinder strength was approximately 3,490 psi. Since the measured strength closely corresponds to the estimated strength, the in-place maturity of 4,000 ?C?hr should correspond to an in- place strength of 4,200 psi. Figure 2.6: Example of estimating strength from strength-maturity relationship (TxDOT 1999) R2 20 The Iowa Department of Transportation specification (2000) for estimating concrete strength using the maturity method will now be discussed and compared to TxDOT?s methods. Iowa DOT also suggests using the Nurse-Saul maturity function with a datum temperature of 14 ?F (-10?C). When developing a strength-maturity relationship, Iowa DOT only requires 12 test specimens and a batch size of 3 yd3. The specimens are moist cured in a saturated sand pit, instead of the common lime-bath or moist-cure room used by TxDOT and recommended by ASTM C 1074. In some cases, Iowa DOT allows the time intervals for temperature recording to be 2-3 hr for the first 24-36 hr and a minimum of twice per day thereafter. Figure 2.7: Example of verification of design strengt h (TxDOT 1999) Iowa DOT requires validation of a strength-maturity curve to be performed once a month on all projects, although it seems that the Iowa DOT uses the maturity method R2 21 mainly for concrete pavement applications. Three test specimens are prepared, and flexural strengths are tested at values as close as possible to the strengths required for the project, such as traffic-opening strengths. The average of the specimen flexural strengths must be within 50 psi of the estimated flexural strength from the strength-maturity curve. If the average flexural strength is more than 50 psi less than the estimated value, a new curve must be developed. If the average flexural strength is greater than 50 psi above the estimated value, a new curve is not required, but may be desired. Figure 2.8: Example of verification of safety or formwork strength As explained in this section, use of the maturity method to estimate the strength of concrete members requires establishment of a unique strength-maturity relationship. Also, some verification processes must be implemented. The maturity method is not 4000 8000 12000 16000 MATURITY INDEX, TTF (?C?HR) COMPRESSIVE STRENGTH (PSI) 6000 5000 4000 3000 2000 1000 Required Cylinder Strength Formwork or safety strength Cylinder maturity Structure maturity 22 perfect however. Users of the maturity method must be aware of certain issues; these limitations are discussed in the next section. 2.3 LIMITATIONS OF THE MATURITY METHOD As noted in the previous section, there are certain factors and natural phenomena that can cause unreliable strength estimations from the maturity method. These include factors that affect concrete strength, such as temperature and mixture proportions, as well as other factors that only apply to the maturity method. This section will focus on all of these such issues. 2.3.1 Mixture -Specific Strength-Maturity Curve In Section 2.2.4 it was stated that the TxDOT requires a new strength-maturity relationship to be established if there are any changes in mixture proportions or major changes in the water-to-cementitious materials ratio. This is because each strength- maturity relationship is unique. In other words, there is no universal maturity curve. Nurse (1949) noticed this in his experiments with temperature and time effects on concrete strength, as previously shown in Figure 2.1. Nurse plotted the maturity index, temperature-time product (?C?hr), on the x-axis and the percentage of 3-day strength on the y-axis. For the two mixtures shown in Figure 2.1, a single curve modeled the strength development well. However, Figure 2.9 shows the same curve plotted with data points from several of Nurse?s other mixtures. It shows that the strength-maturity relationship that fit the two original mixtures well does not model the other mixtures well at all. This 23 occurs because the different mixtures have different long-term strengths and rates of strength gain relative to their unique temperature histories. Figure 2.9: Further results from Nurse?s experiments (1949) Past researchers have attempted to find a single strength-maturity relationship for a certain range of mixtures. Plowman (1958) proposed a single strength-maturity equation based on past studies performed by various authors, with various mixtures, water-to-cement ratios, and curing temperatures between 11?F and 105?F (-12?C to 41?C). Using the logarithmic strength-maturity function (defined in Eq. 2.4), he proposed that the constants, a and b, have specific values based on four strength ranges up to 10,000 psi. With this equation, Plowman found that any concrete strength could be estimated based on a given maturity, regardless of water-to-cement ratio, curing temperature under 100?F, or aggregate-to-cement ratio, with an average error of 3%. Percentage of 3 day Compressive Strength (at 64?F [18 ?C]) Temperature-Time factor (?C?hr) 24 Plowman?s equation seemed to be valid, although his equation was only based on 26 different compressive strength values. The formula did not hold up over time. In a later study, Carino and Lew (1983) examined the effects of temperature on various strength-maturity relationships. Two mortar mixtures with water-to-cement ratios of 0.56 and 0.43 cured at both isothermal and fluctuating temperatures between 41?F and 109?F (5?C to 43?C) were studied. Hyperbolic strength-maturity equations (Eq. 2.5) were fit to each of the mixtures at the various curing temperatures. The best-fit k, Su, and Mo values were determined and plotted versus temperature. The k and Mo values seemed to be independent of water-to-cement ratio; however, the Su plot was much different. Carino and Lew?s results are shown in Figure 2.10. It can be seen that for the same mixture, simply changing the water-to-cement ratio and curing temperature greatly affected the long-term strength of the batches. The high water-to-cement ratio mixture had an average limiting strength of approximately 4,500 psi while the lower water-to- cement ratio mixture had an average limiting strength of approximately 9,500 psi. Recall that the maturity index can only shift time values (see Figure 2.4). Thus, no strength- maturity function can account for strength differences such as those exhibited in Figure 2.10. This effect is uncontrollable and is the reason that specifications call for new strength-maturity curves when there are changes in mixture proportions, admixtures, water-to-cement ratio, and sources of materials. Note that the effect of various curing temperatures on long-term strength can also be seen in Carino and Lew?s results. Figure 2.10 there indicates a general trend: as the curing temperature increases, the limiting strengths decrease. This phenomenon will be discussed further in the next section. 25 Figure 2.10: Ultimate compressive strength (Su) values for Carino and Lew?s experiment (1983) 2.3.2 Effects of Curing Temperature on Long-Term Strength Another major problem that affects the accuracy of strength estimations based on the maturity method is early-age temperature effects on concrete strength. Since Saul?s report on maturity in 1951, it has been well established that early-age curing temperatures may affect the long-term strength of concrete. Saul performed strength tests on rapid- hardening and ordinary concrete mixtures with water-to-cement ratios of 0.35 and 0.50. Saul noted that his maturity rule did not hold if the concrete was subjected to extremely high temperatures at early ages. Specifically, he stated that temperatures above approximately 120?F (50?C) during the first 2 hours and above 212?F (100?C) during the first 6 hours of curing produced the worst effects. 26 The results of Saul?s experiments are shown in Figure 2.11. The solid line on the strength-versus-age graph is the behavior of one of the mixtures ?normally cured? at 62?F (16?C). The dotted lines represent zones of test data for the same concrete mixture subjected to different curing conditions. ?Zone A? represents the strengths of specimens heated to 212?F (100?C) within the first six hours then gradually lowered to approximately 105?F (41?C) within 24 hr. ?Zone B? represents the strengths of specimens heated to 212?F (100?C) within the first six hours, held at that temperature for another eight hours, and then gradually reduced to approximately 130?F (54?C) within 24 hr. The strengths for the specimens represented by ?Zone A? have higher early-age strengths than the ?normally cured concrete? and only have a slight reduction in late-age strengths. The specimens represented by ?Zone B? have slightly higher early-age strengths than the normally cured concrete but have significantly lower strengths at lateages. The 28-day strength for this group of specimens is more than 2,000 psi less than the normally cured specimens in some cases. Figure 2.12 shows the same plot as Figure 2.11, with actual test data instead of the zones. The x-axis has dual units of the Nurse-Saul maturity index and actual age. The data sets represented by dark circles and triangles are the same specimens represented by Zones A and B in Figure 2.11, respectively. The open circle data points represent specimens cured at temperatures between the other two ranges. From Saul?s results, the strengths for the triangle data points, with the worst environmental conditions, depart from the normally cured strength-maturity curve (i.e. the strength-maturity relationship) after less than 2 days. After that, strengths would be greatly over-estimated by the 27 strength-maturity relationship. For the other two data sets, the strength-maturity relationship appears to be accurate only up to approximately 5 days. Figure 2.11: Temperature affects on long-term strength (Saul 1951) In 1956, McIntosh presented results of experiments in which specimens of three mortar mixtures having variable aggregate-to-cement and water-to-cement ratios were cured at temperatures ranging from approximately 59?F to 43?F (15?C to 6?C). McIntosh found that at equal maturities (computed using the Nurse-Saul maturity function), specimens cured at lower temperatures had lower early-age strengths than specimens cured at laboratory temperature (59?F). Moreover, the specimens cured at the lower temperatures had higher 28-day strengths than specimens cured at laboratory temperature, at equal maturities. Thus he concluded that for concrete cured at low Age (days) Compressive Strength (psi) 28 temperatures, the maturity method ?may lead to an over-estimation of the strength at low maturities and an under-estimation at high maturities.? Figure 2.12: Temperature effects on estimating strength using strength-maturity relationship (Saul 1951) Many other researchers such as Klieger (1958), Alexander and Taplin (1962), and Carino and Lew (1983) have verified what Saul and McIntosh found. The results from Alexander and Taplin?s experiment are shown in Figure 2.13. They tested the compressive strength of concrete specimens from a single mixture with a water-to- cement ratio of 0.35 cured at 41?F (5?C), 70?F (21?C), and 108?F (42?C). The x-axis in Figure 2.13 shows maturity given as ?A(T+10).? This is the Nurse-Saul maturity function (Equation 2.1), where A stands for the age of the concrete in days, and the (T+10) term is the temperature term after taking a datum temperature of -10?C (14?F). Nurse-Saul Maturity Index (x 1,000 ?C?hr) Compressive Strength (psi) Actual Age (days) at 62?F (16?C) 29 Figure 2.13: Crossover effect due to curing temperatures (Alexander and Taplin 1962) Alexander and Taplin?s results show exactly what McIntosh and Saul had found. The three data sets have their own unique curves. If the maturity method was valid given any curing history, only one strength-maturity curve would exist for all curing temperatures. However, at early ages, the strength-maturity relationship for the mixture cured at 70?F (21?C) underestimates the strength of specimens cured at higher temperatures and overestimates the strength of specimens cured at lower temperatures. At later ages this trend is reversed. The strength-maturity curve for the mixture cured at 70?F (21?C) underestimates the strength of specimens cured at lower temperatures and overestimates the strength of specimens cured at higher temperatures. Carino (1991) (?C?days) 30 refers to this phenomenon as the ?crossover effect,? as the strength-maturity curves of mixtures cured at colder temperatures literally cross over the strength-maturity curves of mixtures cured at warmer temperatures. In Figure 2.13, the strength-maturity curve for the mixture cured at 41?F (5?C) crosses over the curves for the mixtures cured at 70?F (21?C) and 108?F (42?C) at approximately 375?C?d and 250?C?d, respectively. It could also be said that the strength-maturity curve for the mixture cured at 70?F (21?C) crosses over the curve for the mixture cured at 108?F (42?C) at approximately 150?C?d. Verbeck and Helmuth (1968) investigated the rate of hydration of cement pastes cured at various temperatures in an effort to explain the mechanism responsible for the crossover effect. They proposed that a cement paste cured at high temperatures hydrated at a much higher rate than a paste cured at lower temperatures. The increased rate of hydration does not allow sufficient time for the proper distribution of hydration products, producing a weaker bond. Verbeck and Helmuth concluded that a ?shell? or barrier forms around cement particles of mixtures cured at high temperatures, which hinders further hydration. Thus, concrete that is cured at high temperature gains strength so rapidly that the hydration process is disrupted and the late-age strengths suffer accordingly. The conclusions made by Verbeck and Helmuth have been verified by later researchers through the use of backscattered electron imaging (Kjellsen, Detwiller, and Gj?rv 1991). As seen by results from Saul, McIntosh, and Alexander and Taplin, the crossover effect can lead to severe deficiencies in the accuracy of strength estimations using the maturity method. Much research has been performed on this phenomenon in order to determine if there is a cutoff age beyond which strengths cannot be reliably estimated 31 using the maturity method. From Saul?s study, it seemed that as long as the most severe temperature curing regimens were not used, strengths could be accurately estimated by the strength-maturity relationship up to about 5 days, or approximately 65% of the normally cured mixture?s 28-day strength. Kjellsen and Detwiler (1993) found that the strength-maturity relationships for the mixtures they examined were valid only up to a maturity that corresponded to 40% of the normal 28-day strength. Jonasson (1985) found that strength estimations using the maturity method were valid up to approximately 50% of the normal 28-day strength. To date many researchers have attempted to develop methods of estimating the long-term strength of concrete using modified and more complex versions of the maturity method (Kjellsen and Detwiler 1993; Tank and Carino 1991), but none have been widely accepted in practice. 2.3.3 Other Factors Affecting Concrete Strength There are other factors that affect the strength of concrete but not its heat of hydration, and thus affect strength estimations using the maturity method. Some of these factors include air entrainment, clay particles in aggregates, and inadequate moisture for curing. Air Entrainment It is well known in concrete technology that the more air voids that are in a certain concrete mixture, the lower the strength will be. A general rule is that for every 1% increase in air content, a 5% decrease in strength can be expected (Mindess, Young, and Darwin 2003). This strength loss can be seen in Figure 2.14. In an experiment 32 performed by Cordon (1946), the 28-day compressive strengths of three mixtures with different cement contents and various air contents were examined. From the figure, a general trend can be seen: as the air content increases, the 28-day strength decreases. The mixture with 613 lb/yd3 of cement loses almost 2,000 psi going from 0% to 9% air content. Therefore, if air content is not carefully monitored and controlled then the mixture-specific strength-maturity relationship established becomes inaccurate?as do any resulting strength estimates. Figure 2.14: Effect of air content on concrete strength (Cordon 1946) Clay Fines in Aggregates Another important factor that affects the strength of concrete is the presence of clay fines in aggregates used in any concrete mixture. Top?u and Ugurlu (2003) 613 lb/yd3 519 lb/yd3 425 lb/yd3 Air Content (%) Compressive Strength (MPa) Compressive Strength (1000 psi) 33 examined the affect of very fine aggregates (those passing a No. 100 sieve) on concrete properties. According to Top?u and Ugurlu, clays have a very detrimental affect on the bond between cement and aggregates. They also found that clay slows the hydration process of the cement particles and can change the effective water-to-cement ratio. A simple solution to this problem is to wash aggregates before use. However, it is easy to see how this contamination could influence the predictions made by the maturity method by both affecting strength levels and the rate of hardening of a concrete. Inadequate Moisture for Curing In addition to air voids and fine particles in aggregates, a much more common and debilitating problem affecting the maturity method is moisture (or lack thereof) during curing. The maturity method uses the temperature history to predict the in-place strength of the concrete, but an adequate amount of moisture must be supplied to sustain the hydration while the concrete is gaining strength (ASTM C 1074 2004; Carino 1991). Moisture is a crucial part of the hydration process of concrete. The hardening process of concrete occurs from a ?chemical reaction between the cement and water, called hydration (Kosmatka, Kerkhoff, and Panarese 2002).? The hydration process begins with water supplied in the concrete mixture, but as the chemical reaction continues additional moisture is required. Concrete can continue to hydrate even when it is not fully saturated; however, if the relative humidity drops below approximately 80%, hydration will cease (Kosmatka, Kerkhoff, and Panarese 2002). Therefore, proper moist- curing of concrete is a crucial component in the construction process. Figure 2.15 shows the results of an experiment performed by Gonnerman and Shuman (1928) examining the effects of moist-curing time on the strength gain of concrete. The graph shows that the 34 shorter the moist-cure time, the lower the late-age strength. The strength difference between the concrete moist cured for the entire year and the concrete cured in laboratory air for the entire year is almost 4,000 psi. Thus, if concrete is not cured properly, then the established strength-maturity relationship is no longer valid. The research presented in this section indicates that all of the factors presented must be acknowledged and carefully monitored in order to obtain accurate strength estimations using the maturity method. Figure 2.15: Effect of moist-curing time on strength gain of concrete (Gonnerman and Shuman 1928) 2.4 COMPARISON OF MATURITY FUNCTIONS At this point, both maturity functions and their application have been presented but several questions remain: Age at test (days) Compressive Strength (MPa) Compressive Strength (1000 psi) Moist-cured entire time In air: Entire time After 7 days moist cure After 28 days moist cure 35 ? Does one maturity function model concrete behavior at different curing temperatures more accurately than the other? ? How does the datum temperature or activation energy relate to the accuracy of the strength estimate? ? What datum temperature or activation energy should be used for a particular mixture? This section addresses these remaining issues. First, the mechanism of converting temperature histories to maturity will be examined for both functions. A term, known as the ?age conversion factor? will be presented that will enable comparison of the two functions. Then, the two functions will be compared using actual concrete behavior from past experiments to see which function is believed to best model concrete behavior. Finally, previously determined datum temperatures and activation energies for different mixtures will be presented. 2.4.1 Maturity Function Trends Compared to Concrete Behavior According to Carino (1991), Saul?s maturity rule was based on ?empirical evidence? and assumes specific concrete behavior that may have been erroneous. Later, Freisleben Hansen and Pedersen (1977) proposed the Arrhenius maturity function, a model based on scientific theory of the rates of chemical reactions (Carino 1997). Recall in Figure 2.4 it was shown that the strength versus age data for three mixtures cured at a hot, cold, and control temperature can be converted to one strength-maturity curve. The amount of correction needed to converge the data onto one curve is dependent on how a particular concrete mixture is affected by different curing temperatures, in other words, 36 how the rate of reaction (strength gain) differs given different curing temperatures. However, Nurse-Saul and Arrhenius maturity functions convert temperature histories into maturity using different models. To determine which function estimates strength with the greatest accuracy over a range of temperatures, the approach used in each function to convert temperature histories to maturity must be examined. To do this, the concept of the age conversion factor must be presented. The two maturity functions previously defined in this chapter may be easily compared when the Nurse-Saul maturity function is converted to produce equivalent age, instead of the Temperature-Time Factor. The Nurse-Saul maturity function, when converted to produce equivalent age, can be defined as follows (Carino 1991): Nurse-Saul equivalent age function: ( ) ( ) tTT TTt or oct e ?? ??= 0 Equation 2.6 where, te = equivalent age at the reference curing temperature (hr), Tc = average concrete temperature during the time interval, Dt, (?F or ?C), To = datum temperature (?F or ?C), Tr = reference temperature, (usually either 68?F or 73?F [20?C or 23?C]). When this is done, the Nurse-Saul maturity function converts a curing time to an equivalent age at the reference temperature, analogous to the Arrhenius maturity 37 function. Now both maturity functions, the Nurse-Saul and Arrhenius, are in comparable forms. They both compute equivalent age, and both have some temperature converting function multiplied by ?t. The temperature converting function is known as the ?age conversion factor,? and the equations can now be written in similar form (Carino 1991): Equivalent Age Function: tt te ??= a 0 Equation 2.7 where, ? = age conversion factor, ( )( ) or oc TT TT ? ? , for Nurse-Saul Maturity (NSM) function, ? = age conversion factor, ?? ? ?? ? +?+ ? rc TTR E e 273 1 273 1 , for Arrhenius Maturity (AM) function Now the two maturity functions can be easily compared. The age conversion factor functions can be plotted versus temperature, given certain datum temperatures or activation energies. Some of these plots are presented in Figure 2.16 for a reference temperature of 73?F (23?C). In this figure, the NSM function?s age conversion factors are plotted for datum temperatures of 32?F (0?C) and 14?F (-10?C) and the AM function?s age conversion factors are plotted for activation energies of 40,000 J/mol and 30,000 J/mol. From this graph a distinct difference in the two function?s age conversion factors can be seen. The AM function?s age conversion functions have a non-linear trend versus curing temperature. The NSM function?s age conversion functions have a linear 38 trend versus curing temperature, which is to be expected from their respective formulas. The manner in which these functions fit the actual behavior of concrete at different temperatures is closely related to the accuracy of the maturity method when estimating the concrete strength at different temperatures (Carino 1991). In other words, the age conversion factor is the term that defines the temperature sensitivity for the maturity method. From Figure 2.16, it can be seen that different values of datum temperatures and activation energies produce much different age conversion factors over the temperature range shown. For example, given an average curing temperature of 92?F (33?C), the age conversion factor for the Nurse-Saul maturity function, with a datum temperature of 14?F (-10?C), is approximately 1.3, while the age conversion factor for the Arrhenius maturity function, using an activation energy of 40,000 J/mol, is approximately 1.6. The difference between the two age conversion factors at that temperature is approximately 20% and the difference gets larger as the temperature increases. Thus, using different maturity functions and different datum temperatures or activation energies can cause great deviations in the applied age conversion factor. 39 0.0 0.5 1.0 1.5 2.0 2.5 3.0 32 42 52 62 72 82 92 102 112 122 Concrete Curing Temperature, Tc (?F) Age Conversion Factor, ? Figure 2.16: Age conversion factors for Nurse-Saul and Arrhenius maturity functions To determine which age conversion factor, and thus which maturity function, models concrete strength gain behavior better, Carino (1991) performed a study to examine concrete strength gain behavior at different temperatures. Carino tested the strengths of two mortar mixtures with water-to-cement ratios of 0.43 and 0.56 cured at the following isothermal temperatures: 41?F (5?C), 54?F (12?C), 73?F (23?C), 90?F (32?C), and 109?F (43?C). The strength results were fitted by regression analysis to a hyperbolic strength-age function, which has the same form as Equation 2.5, except time is the independent variable instead of maturity. For clarity, it is shown as follows (Carino 1991): Hyperbolic Strength-Age Function: ? for AM function with: E = 40,000 J/mol E = 30,000 J/mol ? for NSM function with: To = 14?F To = 32?F 40 )(1 )( o o u ttk ttkSS ?+ ?= Equation 2.8 where, S = compressive strength at time, t, (psi), Su = limiting compressive strength (psi), t = time interval (hour), to = time when strength development is assumed to begin (hour), and k = rate constant, initial slope of strength-age curve, (1/hour). The Su, to, and k-values were found from regression analysis, and the k-values were plotted versus curing temperature, since the k-value represents the initial rate of strength gain at the different temperatures. Figures 2.17 and 2.18 were created from Carino?s reported k-values. In Figure 2.17, the k-values are plotted versus temperature, and a linear trendline was fitted. The equation of the line and the R2 value are shown on the plot. Where this trendline crosses the x-axis, the rate constant (k) is zero. According to Carino, that is the datum temperature, to be used with the Nurse-Saul maturity function, that best models the behavior of this concrete mixture. The datum temperature for the mixture used in his experiment, according to the preceding logic, should be 40?F (4.4?C). However, the line does not fit the data very well, which is apparent from the R2 value of 0.93. 41 y = 0.0128x - 0.5112 R2 = 0.9347 0.0 0.2 0.4 0.6 0.8 1.0 1.2 32 44 56 68 80 92 104 116 Curing Temperature (?F) Rate Constant, k (1/day) w/c = 0.43 w/c = 0.56 Figure 2.17: Rate constants used to find best-fit datum temperature (Carino 1991) Next, according to Carino, the activation energy, to be used with the Arrhenius maturity function that best models the strength gain behavior for the mixture studied, may be obtained by plotting the natural logarithm of the k-values versus the inverse of the curing temperature, in Kelvin units. This plot is shown in Figure 2.18. From this figure, it can be seen that the R2 value is 0.99, a much better fit than Figure 2.17. Through mathematical reasoning, Carino showed that the best-fit activation energy to be used for this mixture is the negative of the slope of the trend line multiplied by the universal gas constant. From his results, the activation energy should be 43,700 J/mol. 42 y = -5255x + 16.646 R2 = 0.9858 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 0.0037 Temperature-1 (1/K) Ln(Rate Constant) w/c = 0.43 w/c = 0.56 Figure 2.18: Rate constants used to find best-fit activation energy (Carino 1991) Now, to analyze which maturity function best fits the rate of strength gain of the concrete evaluated in Carino?s experiment, the k-values need to be converted into age conversion factors. Then the age conversion factors from his experiment can be plotted on a graph such as the one shown in Figure 2.16. According to Carino, the rate constant values can be converted into age conversion factors by dividing each of the k-values by the k-value at the reference temperature. The value of the rate constant at the reference temperature, 73?F (23?C), can be found from the trendline given in Figure 2.18. This value comes out to be approximately 0.326 day-1, after transforming from logarithmic values. Next, the newly converted age conversion factors for Carino?s mixture were plotted versus curing temperature in Figure 2.19. The age conversion factor functions for 43 the Nurse-Saul and Arrhenius maturity functions, with the best-fit values of datum temperature or activation energy found previously, were added to the graph as well. From Carino?s findings, the Arrhenius maturity function fits the behavior of the examined mortar much better than the Nurse-Saul maturity function. The non-linear behavior of the mortar used for his study led Carino to conclude that the Arrhenius maturity function was much more suited to model the mortar?s behavior over the curing temperature range of 41?F to 109?F (5?C to 43?C). 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 32 44 56 68 80 92 104 116 Curing Temperature (?F) Age Conversion Factor, ? w/c = 0.43 w/c = 0.56 Figure 2.19: Age conversion factors with best-fit Nurse-Saul and Arrhenius age conversion factors (Carino 1991) However, the age conversion factor for the Nurse-Saul maturity function with datum temperature of 40?F was always less than or equal to than the corresponding age conversion factors for the mortar. Therefore, Carino recognized that the Nurse-Saul ? for NSM function with To = 40?F (4.4?C) ? for AM function with E = 43,700 J/mol 44 maturity function would produce conservative results in this case. In other words, the Nurse-Saul maturity function would underestimate the actual age conversion factor that the concrete required at various temperatures and thus would underestimate the equivalent age. Therefore, strength estimated from the maturity method using an underestimated equivalent age would be conservative as the strengths would be underestimated. Carino is not alone in his conclusions. Many other researchers believe the non- linear Arrhenius maturity function is superior to the linear Nurse-Saul maturity function (Freiesleben Hansen and Pedersen 1977; Guo 1989; Tank and Carino 1991). However, Carino noted the simplicity of the Nurse-Saul maturity functio n and concluded that if it were to be used on the mixture proposed in his study, a dual-datum temperature scheme should be implemented: one datum temperature for concrete cured below the reference temperature and one datum temperature for concrete cured above the reference temperature. From the data already presented in this section, Carino found a value of 36?F (2.2?C) should be used for concrete cured below 73?F and a value of 55?F (12.7?C) should be used for concrete cured above 73?F. The resulting function is shown in Figure 2.20. The function fits the concrete behavior much better over the entire curing range. However, it does overestimate the age conversion factor slightly around 90?F. 45 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 32 44 56 68 80 92 104 116 Curing Temperature (?F) Age Conversion Factor, ? w/c = 0.43 w/c = 0.56 Figure 2.20: Two datum temperature scheme for Nurse-Saul maturity function (Carino 1991) Now that the concept of age conversion factors has been presented and the two maturity functions have been compared, the next section discusses other published values of activation energies and datum temperatures for va rious mixtures proposed by various researchers. Incidentally, ASTM C 1074 outlines a procedure for determining datum temperatures and activation energies for any mixture in Annex A1. This procedure follows the same steps as Carino (1991) followed in his experiment explained above. 2.4.2 Temperature Sensitivity Values As explained in the previous section, datum temperatures and activation energies greatly affect the accuracy of concrete strength estimated by the maturity method. The ? for NSM function with To = 55?F when Tc = 73?F, To = 36?F when Tc < 73?F ? for AM function with E = 43,700 J/mol 46 datum temperature and activation energy will be referred to in this report as ?temperature sensitivity? values of a particular concrete mixture, because they change according to the manner in which the rate of reaction (strength development) of a concrete varies with temperature. In other words, the datum temperature and activation energy found in Carino?s experiment (discussed previously) are best suited for only the mixtures he evaluated. Changing the mixture proportions or adding SCMs would change the strength gain behavior (as explained in Section 2.3.1) and temperature sensitivity. Temperature sensitivity values can be: (1) calculated experimentally, as shown in Section 2.4.1, (2) obtained from equations (Freiesleben Hansen and Pedersen 1977), or (3) approximated from accepted values (ASTM C 1074; TxDOT 1999; Iowa DOT 2000). In this section, some published values of temperature sensitivity for various mixtures with different supplementary cementing materials, water-to-cement ratios, and proportions are reported. 2.4.2.1 Datum Temperatures The datum temperature, to be used with the Nurse-Saul maturity function, represents the temperature below which a particular concrete ceases to gain strength (Bergstrom 1953). In Saul?s report on maturity (1951), he suggested that a temperature of 32?F (0?C) was the temperature below which concrete would cease to gain strength. However, Saul recognized that if concrete was able to set at temperatures above freezing, then cured in temperatures below 32?F (0?C), the datum temperature could be as low as 14?F (-10?C). Later, Plowman conducted a study to investigate at what temperature concrete ceases to gain strength and found 11?F (-12?C) to be the value. ASTM C 1074 (2004) recommends a datum temperature of 32?F (0?C) for use with a Type I cement when no admixtures are used. Should any other cement types or admixtures be used, 47 ASTM recommends the use of the experimental procedure outlined in Annex A1 of ASTM C 1074 (explained in Section 2.4.1 of this thesis). Carino and Tank (1992) conducted an extensive study of the isothermal strength development of concrete and mortar specimens made with different cementitious systems with two water-to-cement (w/c) ratios. Specimens in this study were cured at isothermal temperatures of 50, 73, and 104?F (10, 23, and 40?C), and strength tests were performed at regular age intervals. Carino and Tank obtained the best-fit datum temperatures to be used with the different mixtures using the same procedure as described in Section 2.4.1. Table 2.1 summarizes the datum temperature values obtained from their tests. Carino and Tank concluded that different mixtures called for different datum temperatures and that ?none of the values were -10?C [14?F], which is the value of [To] in the traditional Nurse-Saul maturity function.? Also, it is worth noting that most values found were greater than 32?F (0?C). Table 2.1: Datum temperature values proposed by Carino and Tank (1992) Datum Temperature, To ( F ( C)) Cement Type w/c = 0.45 w/c = 0.60 Type I Type II Type III Type I + 20% Fly Ash Type I + 50% Slag Type I + Accelerator Type I + Retarder 52 (11) 48 (9) 45 (7) 23 (-5) 46 (8) 46 (8) 41 (5) 48 (9) 43 (6) 45 (7) 32 (0) 50 (10) 48 (9) 41 (5) 48 From Table 2.1, the datum temperatures increased with a decrease in the water-to- cement ratio for the Type I and Type II mixtures. The datum temperature for the Type III mixture was not affected by the change in water-to-cement ratio. Their Type I + 20% Fly Ash mixture was the Type I mixture with 20% of the cement replaced by fly ash. This mixture required the lowest datum temperature of 23?F (-5?C) at a water-to-cement ratio of 0.45. The Type I + 50% Slag mixture was the Type I mixture with 50% of the cement replaced by GGBF Slag. The datum temperature for that mixture increased with an increase in the water-to-cement ratio. Carino and Tank also experimented with the effect of adding chemical admixtures such as accelerator and retarder to the Type I mixture. The Type I + Accelerator mixture required a datum temperature of 46?F (8?C) and the Type I + Retarder mixture required a datum temperature of 41?F (5?C) for the lower water-to-cement ratio mixtures. The addition of chemical admixtures thus changed the datum temperature of the concrete. The average of Carino and Tank?s results for the lower water-to-cement ratio mixtures, minus the Type I + 20% Fly Ash mixture, was approximately 46?F (8?C). These results, as well as those from Carino, Bergstrom, and Saul show that different mixtures can require very different datum temperatures for use with the Nurse-Saul maturity function. 2.4.2.2 Activation Energies The concept of an activation energy came from Svante Arrhenius in 1888 and accounts for the extra energy required to begin a chemical reaction (Carino 1997). Carino gives an analogy of this process as the energy required to push a brick, standing upright, just enough to start the brick in motion, where gravity will control the remaining 49 process. The initial push would be the activation energy in that analogy. Different mixtures require different energies to begin the hardening process. ASTM C 1074 (2004) recommends activation energies in the range of 40,000 to 45,000 J/mol for use with a Type I cement with no admixtures. Should any other cement types or admixtures be used, ASTM recommends the experimental procedure outlined in Section 2.4.1. Freiesleben Hansen and Pedersen (1977) developed an equation to compute activation energies based on the concrete?s curing temperature. The Freiesleben Hansen and Pedersen (FHP) equation is as follows: FHP Model for Computing the Activation Energy: (Equation 2.9) for Tc ? 20?C (68?F) E = 33,500 J/mol, and for Tc < 20?C (68?F) E = 33,500 + 1,470 (20-Tc) J/mol. This model was determined by computer analysis of early-age compressive strength data, where it was found that concretes cured at hot temperatures required a lower activation energy, in this case 33,500 J/mol. However, as curing temperatures dropped below the reference temperature, a much higher activation energy was needed. For example, a concrete cured at 41?F (5?C) would require an activation energy of approximately 56,000 J/mol, using the FHP model. To compare this equation to constant values of activation energy, Figure 2.21 shows the age conversion factors for the Arrhenius maturity function using the FHP model, as well as using the constant values of 40,000 J/mol and 30,000 J/mol. The Nurse-Saul maturity function is also plotted with a value of 32?F (0?C) as the datum temperature. The reference temperature for all of the curves in this figure is 68?F (20?C), to be consistent with the FHP model. The figure shows that at lower temperatures, the age conversion factors from the FHP model are less 50 than those from E = 40,000 J/mol. At warmer temperatures the age conversion factors determined from the FHP model fall in between those determined with E = 30,000 J/mol and 40,000 J/mol. Note that the age conversion factor for the Nurse-Saul maturity function using 32?F (0?C) as the datum temperature is close to the FHP model at temperatures below the reference temperature. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 32 42 52 62 72 82 92 102 112 122 Concrete Curing Temperature, Tc (?F) Age Conversion Factor, ? Figure 2.21: Comparing age conversion factors using FHP model to other functions There have been numerous researchers who have determined and published their own activation energies for various mixtures. A list of some typical published values of activation energies, reported by Carino (1991), is given in Table 2.2. From the table, the Type I mixtures Carino reported required activation energies from 41,000 J/mol to 44,000 J/mol, which agrees with the values recommended by ASTM C 1074. The remaining values are the activation energies found from the various types of testing, such as heat of ? for AM function with, E = 40,000 J/mol E = FHP model E = 30,000 J/mol ? for NSM function with, To = 32?F 51 hydration and chemical shrinkage, and various mixture designs. The table shows elevated values of activation energy of 49,000 and 56,000 J/mol for mixtures with slag replacement doses of 50% and 70%, respectively. Table 2.2: Activation energy values proposed by various research efforts (Carino 1991) Cement Type Type of Test Activation Energy, E (J/mol) Type I (Mortar) Type I (Mortar) Type I (Concrete) OPC* (Paste) OPC* + 70% Slag OPC* (Paste) RHC* (Paste) OPC*(Paste) Type I/II (Paste) Type I/II + 50%Slag (Paste) Strength Strength Strength Heat of Hydration Heat of Hydration Chemical Shrinkage Chemical Shrinkage Chemical Shrinkage Heat of Hydration Heat of Hydration 42,000 44,000 41,000 42,000 ? 47,000 56,000 61,000 57,000 67,000 44,000 49,000 *OPC = ordinary portland cement. RHC = rapid hardening cement. The RILEM Technical Committee 119 - TCE recommends using the FHP model for all portland cements (Springenschmid 1998). However, for slag cement mixtures cured at any temperature, RILEM suggests using approximately 50,000 J/mol. Carino and Tank (1992) also obtained the activation energy values for their experiment with different mixtures cured at various isothermal temperatures, as described above. Table 2.3 summarizes the activation energies found from their tests. From this table it can be seen that for the Type III and Type I + Retarder mixtures, the change in the water-to-cement ratio had no effect on the activation energies. However, with Type I and 52 Type II cements, the lower water-to-cement ratio mixtures had significantly higher activation energies. On the other hand, the Type I + 50% GGBF Slag mixtures had higher values of activation energy as the water-to-cement ratio increased. These trends are the same as those found for the datum temperatures, described earlier. Carino and Tank?s results further indicate how changes to mixture proportions can alter the activation energy for a particular concrete mixture. Table 2.3: Activation energy values proposed by Carino and Tank (1992) Activation Energy, E (J/mol) Cement Type w/c = 0.45 w/c = 0.60 Type I Type II Type III Type I + 20% Fly Ash Type I + 50% Slag Type I + Accelerator Type I + Retarder 64,000 51,000 44,000 30,000 45,000 45,000 39,000 48,000 43,000 44,000 31,000 56,000 50,000 39,000 2.6 SUMMARY As shown in this chapter, the maturity method can be used to estimate concrete strengths at various curing temperatures using either the Nurse-Saul or Arrhenius maturity functions with a mixture-specific strength-maturity relationship. The mixture- specific strength-maturity relationship must be predetermined prior to any strength estimations. Measures must be taken to verify the estimated in-place strength. Quality control of batching and raw materials and construction techniques must be maintained in 53 order for the predetermined strength-maturity relationship to remain appropriate for that mixture. Long-term strength loss due to high curing temperatures, or crossover, must be taken into account as well. A comparison of the two popular maturity functions was also given in this chapter. It sho uld be reemphasized that the Nurse-Saul maturity (NSM) function has a linear age conversion factor function and the Arrhenius maturity (AM) function has a non-linear age conversion factor function. These trends were essential in determining the best matur ity function to use to estimate concrete strengths of the mixtures in this study. It is possible that a linear age conversion factor function could be well suited for some mixtures, or types of mixtures, while a non-linear age conversion factor function may be better suited for others. The method for determining the mixture-specific datum temperatures and activation energies recommended by ASTM C 1074 (2004), described in Section 2.4.1, was used extensively in the portion of this study described in Chapter 6. 53 CHAPTER 3 LABORATORY TESTING PROGRAM This chapter provides details of the experimental program developed to evaluate the accuracy of using the maturity method to estimate the strength development of hardening concrete cured under various temperature histories. This includes description of the concrete mixtures chosen to be evaluated as well as the raw materials, proportions, and curing methods used. 3.1 CONCRETE MIXTURES CONSIDERED Thirteen concrete mixtures were evaluated in this study. A summary of the mixtures can be found in Table 3.1. The mixtures were selected in order to evaluate the effect of different dosages of supplementary cementing materials (SCMs), different SCM types, varying cement types, and changes in the water-to-cementitious ma terials (w/cm) ratio. The mixtures were chosen because they are similar to common mixtures used for ALDOT construction projects. For example, there are normal strength mixtures, bridge deck mixtures, repair mixtures for higher early strength, and prestressed concrete mixtures. The Type I HPC Bridge Deck mixture with a w/cm ratio of 0.41 can be compared to the Class C fly ash, Class F fly ash, and the GGBF Slag mixtures, since they all have the same mixture proportions, except that the SCM is used to replace the specified percentage of Type I cement. This percentage is based on mass of cement 54 replaced by SCMs. For example, the 20% C mixture is the same as the Type I - 0.41 mixture, except with 20% of the cement by mass replaced with Class C fly ash. These mixtures allow the effect of various SCMs on concrete strength when cured under different temperatures to be examined. Table 3.1: Types of mixtures evaluated Mixture Identification Cementitious Materials Content w/cm Classification CEMENT ONLY Type I - 0.48 620 lb/yd3 0.48 Normal: A-1a Type I - 0.44 620 lb/yd3 0.44 Normal: A-1c Type I - 0.41 658 lb/yd3 0.41 HPC Bridge Deck Type III - 0.44 620 lb/yd3 0.44 Repair: A-1c Type III - 0.37 705 lb/yd3 0.37 Prestressed Girder CLASS F FLY ASH 20% F 0.41 Bridge Deck 30% F 658 lb/yd3 0.41 Bridge Deck CLASS C FLY ASH 20% C 0.41 Bridge Deck 30% C 658 lb/yd3 0.41 Bridge Deck GGBF SLAG 30% Slag 0.41 Bridge Deck 50% Slag 658 lb/yd3 0.41 Bridge Deck SILICA FUME 70/20/10 - 0.44 620 lb/yd3 0.44 HPC Bridge Deck: A-1c 70/20/10 - 0.37 705 lb/yd3 0.37 Prestressed girder The mixtures with 620 lb/yd3 of cement or cementitious material originate from the standard ALDOT mixtures, A-1c and A-1a, defined in Section 501.02 (ALDOT 55 2002). The A-1c mixture was modified by the type of cement and replacement of cement with SCMs. The 70/20/10 mixtures represent mixtures with cementitious material content made up of 70% Type I cement, 20% Class F fly ash, and 10% silica fume. These types of mixtures are commonly used by ALDOT in structures exposed to a marine environment. Mixures with SCMs require the use of a water-to-cementitious material (w/cm) ratio, as opposed to the water-to-cement ratio. This is done to show that, although the water-to-cement ratio changes from the straight cement mixtures to the mixtures with SCMs, the water-to-cementitious materials ratio stays constant. 3.2 RESEARCH APPROACH All thirteen mixtures were batched at different temperatures and then cured under three different temperature histories in order to evaluate the maturity method over the full practical temperature range. The following curing conditions were used: 1. One batch of each mixture was mixed and cured at a constant ?control? temperature (i.e. 68 - 73?F). 2. One batch of each mixture was batched and cured under a representative ?hot? temperature history to simulate summer conditions. 3. One batch of each mixture was batched and cured under a representative ?cold? temperature history to simulate winter conditions. Both the ?hot? and ?cold? batches were cured under fluctuating temperatures, which simulate actual field conditions more realistically than isothermal conditions. These fluctuating curing conditions were achieved by constructing two insulated water tanks. 56 The water temperature in each tank was controlled by half-inch copper pipes that oscillated through the tank. The temperature of the water in the copper pipes was controlled by a programmable, heating and cooling water circulator. Figures 3.1 and 3.2 depict one of the tanks and circulator used to create the hot and cold curing temperatures. The tanks were approximately 8 x 2.5 x 2 ft surrounded by a wood frame that held 4-in.- thick closed-cell polystyrene insulation sheets. Each 1-ft3 capacity programmable circulator (Polyscience, Model No. 9612) was connected to a circuit of ?-in. copper pipes, which transferred heat from the water in the circulator to the lime-saturated water in the appropriate tank. The tanks were designed to hold approximately forty 6x12 in. cylinders. From each batch of concrete, nineteen 6x12 in. cylinders were prepared and tested at various ages, in order to evaluate the strength gain behavior of the different mixtures. The tanks were designed to hold the fresh cylinders, without being fully submerged, as well as hold any late-age cylinders from different mixtures, yet to be tested. The cylinder testing plan is explained further in Chapter 4. Figure 3.1: Curing tank and programmable heating/cooling circulator 57 The temperatures in the circulators had to be calibrated to achieve the desired temperatures in the lime-saturated water baths. The desired curing temperatures for the three batches of each mixture are shown in Figure 3.3. The curing temperature cycle in the lime-saturated water for the ?hot? mixtures was 90?F to 106?F and back to 90?F in 24 hours. This range was cycled for the entire curing period. For the setup in this project, the ?hot? circulator had to be programmed to run from 85?F to 120?F and back to 85?F in 24 hours to achieve the desired temperatures in the lime-bath. The curing temperature cycle in the lime-saturated water for the ?cold? mixtures was from 55?F to 40?F and back to 55?F in 24 hours. For the setup in this project, the ?cold? circulator was programmed to run between 28?F and 65?F. A mixture of approximately 30% to 80% of ethylene glycol to water was used in the cold circulator to prevent freezing. The curing temperature ranges were chosen in order to fully cover the range of practical temperatures found on most ALDOT construction projects. The cold temperatures were chosen to meet the lowest temperature (i.e. 40?F) that concrete can be poured, according to ALDOT standard specifications (ALDOT 2002). The hot temperatures were chosen to show the effects of rapid hardening on strength-gain behavior of concrete and the resulting effects on strength estimations using the maturity method. Once compressive strength testing was completed for each batch, the temperature history of the cylinders was obtained, and the data were converted to maturity values. More details of the testing procedures and methods are reported in Chapter 4. 58 Figure 3.2: Inside of a temperature-controlled curing tank Time (hours) Cylinders placed in tank 106 90 73 68 55 40 Control Curing Temperatures Cold Curing Temperatures 24 48 72 96 120 Hot Curing Temperatures Curing Tank Temperature ( ?F) 59 Figure 3.3: Curing tank cycles 3.3 RAW MATERIALS 3.3.1 Cements and Supplementary Cementing Materials All Type I Cement was produced by LaFarge from their Roberta plant in Calera, Alabama. Due to the length of the project, four different orders of Type I cement were used. Although these cements came from the same producer, they were treated as coming from separate sources. All bridge deck mixtures with 658 lb/yd3 of cementitious material content contained the Type I cement labeled Source A. The mixtures Type I - 0.44 and Type I - 0.48 contained the Type I cement named Source B. Source C was used for the Type I - 0.44 control batch. It came from a separate source because a data collection failure required the mixture to be reproduced. This was considered acceptable, as the two cements are nearly identical (see Table 3.2). Source D was used for the ternary mixtures (70/20/10). The Grade 120 ground-granulated blast-furnace slag was obtained from Buzzi Unicem USA, Inc. from their New Orleans Slag Facility. The Class C fly ash was obtained from Holcim US, Inc. from their plant in Quinton, AL, and only one source was used. Two sources of Class F fly ash were used in this project: one was used for the Class F fly ash bridge deck mixtures, and the other was used for the 70/20/10 mixtures in combination with Type I cement and silica fume. They are labeled Class F1 and F2 for the Class F fly ash bridge deck and 70/20/10 mixtures, respectively. Both sources of Class F fly ash were produced by Boral Material Technologies from their Plant Bowen, near Stilesboro, Georgia. The densified Silica Fume was produced by Simcala, Inc. located in Mt. Meigs, Alabama. The Type III mixtures used cement that was produced by Cemex out of their Demopolis, Alabama plant. 60 All of the cementitious materials used in this study were sent to Wyoming Analytical Laboratory?s Denver, Colorado branch to be tested for their chemical compositions. In addition, all the cementitious materials were tested for specific surface area, or Blaine, by Mactec, Inc in Atlanta, Georgia. The results of the analyses are summarized in Tables 3.2 and 3.3. All cements met the chemical requirements for ASTM C 150. The slag, Class F fly ash, Class C fly ash, and silica fume were all certified by the specific producers as passing the relevant ASTM specifications. From Table 3.2, it can be seen that the Type I cements, sources A through C, had similar compositions. Their C3S percentages ranged from 67 to 55. The Type I, source D, had some slight variations from sources A, B, and C. Source D had the lowest percentage of CaO and the highest percentage of SiO 2 and Al2O3, which are the largest factors in calculating the percentage of C3S as 48.76. Because the values were not far off and because source D was used only for the ternary mixtures, the deviation was considered acceptable for the purposes of this project. 61 Table 3.2: Chemical composition of cementitious materials Portland Cement Type I Fly Ash Parameter I A IB IC ID Type III Class C Class F 1 Class F2 GGBF Slag Silica Fume Silicon dioxide, SiO 2 (%) 20.32 20.85 20.79 21.31 20.23 39.94 53.09 54.85 32.68 95.57 Aluminum oxide, Al2O3 (%) 4.75 4.47 4.63 4.97 4.90 18.51 29.10 28.34 9.67 0.02 Iron oxide, Fe2O3 (%) 3.03 2.96 2.89 3.07 2.61 5.71 7.54 7.54 1.12 0.05 Calcium oxide, CaO (%) 65.12 64.11 63.14 62.96 63.74 23.01 1.56 1.31 45.32 0.41 Free CaO (%) 0.21 0.24 0.30 0.13 0.24 - - - - - Magnesium oxide, MgO (%) 2.46 2.77 3.30 2.70 1.11 5.26 0.94 0.90 7.40 0.43 Alkalies (Na2O + 0.658K2O) (%) 0.25 0.28 0.28 0.27 0.31 2.25 2.04 2.05 0.32 - Sulfur trioxide, SO3 (%) 2.56 2.90 2.89 2.75 4.45 1.56 0.10 0.03 1.66 0.10 Loss on ignition, (%) 1.10 1.19 1.52 1.46 1.86 0.23 2.19 1.94 0.84 2.43 Tricalcium silicate, C3S (%) 67.08 60.01 55.54 48.76 56.33 - - - - - Dicalcium silicate, C2S (%) 7.66 14.51 17.72 24.32 15.52 - - - - - Tricalcium aluminate, C3A (%) 7.48 6.83 7.36 7.96 8.57 - - - - - Tetracalcium aluminoferrite, C4AF (%) 9.21 9.00 8.81 9.35 7.96 - - - - - 62 Table 3.3: Specific surface areas and specific gravities of all cementitious materials Material Blaine, Specific surface area (m2/kg) Specific gravity Type IA 380 3.15 Type IB 360 3.15 Type IC 380 3.15 Type ID 350 3.15 Type III 490 3.15 Class C fly ash 380 2.63 Class F1 fly ash 230 2.29 Class F2 fly ash 210 2.29 GGBF Slag 550 2.91 Silica Fume 1200 2.30 3.3.2 Aggregates Two sources of both coarse and fine aggregates were used in this study. A No. 67 crushed limestone from Martin Marietta?s O?Neil, Alabama quarry was used with fine aggregate from Martin Marietta?s Shorter, Alabama quarry for all mixtures except the prestressed concrete mixtures. A No. 78 crushed limestone from Vulcan Materials? Calera, Alabama quarry and fine aggregate from Superior Products? Jemison, Alabama quarry were used for the two prestressed concrete mixtures (the mixtures with water-to- cementitious material ratio of 0.37). The gradations can be seen in Figures 3.4 to 3.7. In these figures, the fine aggregates are labeled by their plant locations, and the coarse aggregates are labeled by their aggregate size. All four aggregates satisfied gradation 63 requirements of ASTM C 33 and AASHTO M 43 specifications. The AASHTO (2003) specification was used since ASTM does not have a No. 78 gradation. The bulk specific gravities and absorption values also were obtained for mixture proportioning. The results from these tests, outlined in ASTM C 127 and C 128, are summarized in Table 3.5. 0 10 20 30 40 50 60 70 80 90 100 3/8" No. 4 No. 8 No. 16 No. 30 No. 50 No. 100 Sieve Size Mass Percent Passing Shorter Fine Aggreagate ASTM C33 Lower Limit ASTM C33 Upper Limit Figure 3.4: Gradation test results for Martin Marietta Fine Aggregate 64 0 10 20 30 40 50 60 70 80 90 100 3/8" No. 4 No. 8 No. 16 No. 30 No. 50 No. 100 Sieve Size Mass Percent Passing Jemison Fine Aggregate ASTM C33 Lower Limit ASTM C33 Upper Limit Figure 3.5: Gradation test results for Superior Products Fine Aggregate 0 10 20 30 40 50 60 70 80 90 100 1" 3/4" 3/8" No. 4 No. 8 Sieve Size Mass Percent Passing No. 67 Crushed Limestone ASTM C33 Lower Limit ASTM C33 Upper Limit Figure 3.6: Gradation test results for Martin Marietta No. 67 Limestone 65 0 10 20 30 40 50 60 70 80 90 100 3/4" 1/2" 3/8" No. 4 No. 8 No. 16 Sieve Size Mass Percent Passing No. 78 Crushed Limestone AASHTO M 43-88 Lower Limit AASHTO M 43-88 Upper Limit Figure 3.7: Gradation test results for Vulcan Materials No. 78 Limestone Table 3.4: Specific gravities and absorptions for all materials Aggregate Bulk Specific Gravity Absorption % Martin Marietta?s Fine Aggregate 2.66 0.67 Superior Products? Fine Aggregate 2.63 0.5 No. 67 Crushed Limestone 2.73 0.99 No. 78 Crushed Limestone 2.74 0.30 3.3.3 Chemical Admixtures Chemical admixtures were used as needed in the mixtures to control the slump and air content of the fresh concrete. All chemical admixtures were supplied by Degussa Admixtures, Inc., and recommended dosages were used whenever possible. For all mixtures, except the prestressed concrete mixtures, Pozzolith 200N was used as a water- 66 reducing admixture and MB AE 90 was used as the air-entraining admixture. Dosages are described in Section 3.4. For the prestressed concrete mixtures, where a higher slump was needed because of the lower w/cm ratio, Glenium 3200 HES was used as a high- range water-reducing admixture; Polyheed 1025 was used as a mid-range water-reducing admixture; and Micro Air was used as the air-entraining admixture. The individual chemical admixture specifications may be obtained from the manufacturer. 3.4 MIXTURE PROPORTIONS The mixture proportions for all thirteen mixtures evaluated in this study are given in Table 3.5. Admixture doses used are given in both oz/yd3 and ounces per hundred weight of cementitious ma terial. The water-to-cementitious materials ratio (w/cm) is given as explained in Section 3.1. From Table 3.5, it can be seen that as the water-to- cementitious materials ratio increased for the Type I straight cement mixtures, the dosages of the water-reducing and air-entraining admixtures decreased. Also note that the bridge deck mixtures with cementitious material content of 658 lb/yd3 all have the same mixture proportions except the SCM types and dosages are variable. This caused some variations in the fresh concrete properties, as reported in Chapter 5. However, this was necessary in order to isolate the effects of the various types and doses of SCMs. 67 Table 3.5 (a): Mixture proportions for all concrete mixtures used in this study (Continued?) Mixture Identification Constituent Type I - 0.41 Type I - 0.44 Type I - 0.48 20% F 30% F 20% C 30% C 30% Slag 50% Slag Water (lb/yd3) 267 298 273 267 267 267 267 267 267 Type I Cement (lb/yd3) 658 620 620 526 461 526 461 461 329 Class F Fly Ash (lb/yd3) - - - 132 197 - - - - Class C Fly Ash (lb/yd3) - - - - - 132 197 - - GGBF Slag (lb/yd3) - - - - - - - 197 329 Coarse aggregate (lb/yd3) 1824 1921 1963 1824 1824 1824 1824 1824 1824 Fine aggregate (lb/yd3) 1210 1115 1138 1168 1147 1188 1177 1196 1187 Water-reducing admixture, oz/yd3, (oz/cwt) 19.7 (3.0) 9.3 (1.5) - 19.7 (3.0) 19.7 (3.0) 19.7 (3.0) 19.7 (3.0) 19.7 (3.0) 19.7 (3.0) Mid-range water-reducing admixture, oz/yd3, (oz/cwt) - - - - - - - - - High-range water-reducing admixture, oz/yd3, (oz/cwt) - - - - - - - - - Air-entraining admixture, oz/yd3, (oz/cwt) 3.0 (0.46) 2.0 (0.32) 1.5 (0.24) 3.0 (0.46) 3.0 (0.46) 3.0 (0.46) 3.0 (0.46) 3.0 (0.46) 3.0 (0.46) Water-to-cementitious material ratio (w/cm) 0.41 0.48 0.44 0.41 0.41 0.41 0.41 0.41 0.41 68 Table 3.5 (b): Mixture proportions for all concrete mixtures used in this study Mixture Identification Constituent Type III - 0.37 Type III - 0.44 70/20/10 - 0.37 70/20/10 - 0.44 Water (lb/yd3) 250 286 261 273 Type I Cement (lb/yd3) - - 494 434 Type III Cement (lb/yd3) 705 620 - - Class F Fly Ash (lb/yd3) - - 141 124 Silica Fume (lb/yd3) - - 71 62 Coarse aggregate (lb/yd3) 1944 1947 1885 1926 Fine aggregate (lb/yd3) 1111 1141 1092 1113 Water-reducing admixture, oz/yd3, (oz/cwt) - 17 (2.75) - 21.7 (3.5) Mid-range water-reducing admixture, oz/yd3, (oz/cwt) 28.2 (4.0) - 28.2 (4.0) 21.7 (3.5) High-range water-reducing admixture, oz/yd3, (oz/cwt) 42.3 (6.0) - 42.3 (6.0) - Air-entraining admixture, oz/yd3, (oz/cwt) 21.2 (3.0) 0.6 (0.1) - 2.0 (0.32) Water-to-cementitious material ratio (w/cm) 0.37 0.44 0.37 0.44 69 3.5 SUMMARY In this chapter, the setup of the experiment and materials used were described. A total of thirteen mixtures were batched and cured at three different temperatures. The ranges of curing temperatures were 90?F to 106?F and 55?F to 40?F, for the ?hot? and ?cold? batches, respectively. These batches fluctuated between their respective temperatures on a 24-hour cycle. There was an additional ?control? mixture that was batched and cured under normal laboratory conditions (i.e. 68 - 73?F). The cements and supplementary cementing materials (SCMs) used were defined, and the chemical analyses were given. The aggregates were tested for gradation, specific gravity, and absorption values and the results were shown. The mixture proportions for all thirteen mixtures were also defined. The effect of the various curing temperatures on the compressive strength was evaluated for the following variables: ? different cement types (Type I and III) ? w/cm = 0.37, 0.41, 0.44, and 0.48 ? Supplementary cementing materials type and dosage o Class F fly ash at 20 and 30% replacement levels o Class C fly ash at 20 and 30% replacement levels o Ground-granulated blast-furnace slag at 30 and 50% replacement levels o Silica fume at a 10% replacement level in a ternary blend with 20% Class F fly ash. The testing performed to evaluate the effect of these variables on concrete strength is explained in the next chapter. 70 CHAPTER 4 LABORATORY PROCEDURES AND TESTS The tests and procedures carried out in order to evaluate the accuracy of using the maturity method to estimate the concrete strength development of hardening concrete cured under various temperature histories are discussed in this chapter. This includes the documentation of the concrete production procedures, fresh concrete test procedures, and hardened concrete test procedures. 4.1 CONCRETE PRODUCTION PROCEDURES 4.1.1 Mixing Procedures The concrete mixtures and the respective raw materials used were defined in Chapter 3. In order to achieve the correct proportions of chemical admixtures, trial batches were performed on all thirteen mixture designs. A trial batch was generally proportioned to be 2 ft3. This allowed the mixture to be tested for slump, air content, unit weight, and to check compressive strength levels. Mixtures using the Glenium 3200 admixtures required an increase in trial batch sizes to 5.6 ft3, because of problems scaling the batch size. All trial batches were mixed at room temperature (i.e. 68 - 73?F). The target air content and slump for all mixtures are listed in Table 4.1. These values are similar to those from ALDOT Standard Specifications Section 501.02 (ALDOT 2002). Target air content and slump were based on the mixtures batched at the control temperatures. It was necessary to keep all chemical admixture doses constant within a 71 mixture. Therefore, some temperature effects on slump and air content were unavoidable. Table 4.1: Target slump and air content for all mixtures Mixture Type Slump Air Content All mixtures except prestressed concrete 2 to 5 in. 3 to 6 % Prestressed concrete < 8 in. 3 to 6 % After chemical admixture dosages and the mixture proportions were verified with trial batches, each mixture was mixed at ?hot?, ?cold?, and ?control? temperatures as previously described. The batch size for all full mixtures was 5.6 ft3. Prior to mixing, all materials used for a particular mixture were either cooled or heated in accordance with the desired mixing temperature. This was essential in order to reproduce actual batch- plant procedures, where stockpiles and materials are stored in ambient conditions. All aggregates, cement, supplementary cementing materials (SCMs), and water to be used in a certain batch were placed in an environmental room at the appropriate temperature for at least 2 days. For the hot mixtures, the room was set at approximately 115?F, which produced an average batch temperature of 101?F. The room was set to approximately 40?F for the cold mixtures, which resulted in an average batch temperature of 55?F. Before weighing and batching the coarse aggregate, fine aggregate, and water, the moisture contents of the aggregates were calculated per ASTM C 566. It was important to wait until the mixing day to test moisture content, especially for the hot and cold mixtures, because of the potential moisture variations introduced by storing these 72 materials in the environmental chamber. Before sampling the fine aggregate, it was mixed thoroughly to obtain a representative sample. A sample of about 800 g was weighed and then heated until dry. The sample was weighed and heated until the weight was within approximately 0.1 g of the previous reading. The moisture content was then compared to the absorption capacity of the aggregate, and the batch water was adjusted accordingly. For example, if the moisture content of the fine aggregate was less than the expected value for the saturated, surface-dry state, then the aggregate would absorb some of the batch water, thereby affecting the water-to-cementitious materials ratio. Thus, extra batch water and less fine aggregate would be required in this case to keep the water- to-cementitious materials ratio as designed. Once the aggregate moisture contents were found and accounted for, all the raw materials were weighed and batched to the proper proportions. All mixing followed ASTM C 192 procedures. A ?butter? batch of fine aggregate, water, and cement was used to coat the inside of the 15 ft3 mixer. Generally 1/4 of a 5-gal. bucket of fine aggregate and two large scoops of cement were used. Water was then added to the mixer to obtain the correct consistency of the butter mixture. Care was exercised not to add excessive water, as this would affect the water-to-cementitious materials ratio. Once the mixer was prepared, the butter mixture was discarded. Next, all aggregates were added to start the actual mixing process. To ensure proper mixing, 5-gal buckets of coarse and fine aggregates were alternated as they were added into the mixer. Once all aggregates were in the mixer, 80% of the batch water was added along with any air-entraining chemical admixture. After two minutes of mixing, the cement and any SCMs were added, as well as any water-reducing chemical admixtures. The concrete was then mixed 73 for three minutes, followed by three minutes of rest and a final two minutes of mixing. Next, testing of slump, air content, density, and temperature were performed. 4.1.2 Mixing with Silica Fume Silica fume can cause high air content in concrete mixtures mixed in laboratory settings (Holland 2005). Because the silica fume particles are very small, the manufacturer densifies them into (relatively) larger conglomerates that make the material easier and safer to handle. However, these conglomerates can trap unwanted air in concrete mixtures if the proper care is not taken to separate all silica fume conglomerates. The procedure outlined by the Silica Fume Association was used (Holland 2005). The mixing procedure was the same as described above except that along with the coarse and fine aggregates and 80% water, the silica fume was added and mixed for approximately 5 minutes to break up the silica fume conglomerates. Next, the cement and Class F fly ash were added, and the mixing procedure followed as described above. This procedure was very successful in eliminating entrapped air due to the presence of densified silica fume conglomerates. 4.2 FRESH CONCRETE PROPERTY TESTING Slump, air content, density, and temperature testing were all performed according to ASTM C 143, C 231, C 138, and C 1064, respectively. The author and his colleague performed all tests and they were ACI-certified as ?Field Technician: Level I? before starting any testing for this project. Nineteen 6x12 in. cylinders were made for each batch, as previously explained. Cylinders were prepared according to ASTM C 192. 74 One cylinder was used to measure the temperature and thus the maturity of the concrete. The temperature was measured using a Dallas Semiconductor programmable i-Button?. Because the i-Buttons have a finite memory, they were programmed to read at time intervals appropriate to reach 28-days of equivalent age (this will be exp lained in Section 4.3.1). The intervals were generally 25, 20, and 30 minutes for the control, hot, and cold batches, respectively. After the cylinders were made, they were covered by snap-on plastic lids to prevent moisture loss and transported to the water tanks to begin curing as soon as possible. Whenever possible, the concrete was mixed so that it would be placed in the lime-saturated bath while the curing tanks were heating. This was chosen to simulate a day-time placement on a job-site. As per ASTM C 1074 (2004), the first set of compressive strength tests were performed at twice the time that the concrete took to reach final set. Final set times were obtained from penetration resistance tests according to ASTM C 403 (2004). Mortar was sieved through a No. 4 sieve from the concrete mixed at the three different temperatures. The mortar specimen was placed in the appropriate curing tank to accurately assess the setting behavior of the concrete specimens. At 24 hours of equivalent age, all cylinders were stripped from their plastic molds; three were tested for compressive strength, and the remaining cylinders were returned to the lime-saturated water tank, fully submerged. The other tests were performed at the times specified in the next section. 75 4.3 HARDENED CONCRETE TESTING Eighteen of the 6x12 in. cylinders were used for compressive strength testing. Three cylinders were tested according to AASHTO T 22 (2003) at each time of interest. The tests were carried out at the following equivalent ages: twice the final set time, 24 hr, 48 hr, 7 day, 14 day, and 28 day. Equivalent age was used so that data points on the strength-maturity plots would be at comparable strength levels for batches of the same mixture. The actual test times for the cold and hot batches were obtained by using the Nurse-Saul equivalent age formula (Equation 2.6) with a datum temperature of 14?F (-10?C). The average of the desired curing temperature ranges (given in Section 3.2) of the cold and hot batches were used: 47.5?F (9?C) and 98?F (37?C) for the cold and hot batches, respectively. The reference temperature used for these equivalent age calculations was 73?F (23?C). The given equivalent ages were approximately equal to the actual test times for the control batches because the control batches were cured so close to the reference temperature. To find the actual test times for the hot and cold mixtures, the age conversion factors (defined by Equation 2.7) for the respective curing temperatures were computed. The age conversion factors for the cold and hot mixtures were 0.57 and 1.42, respectively. To find the actual test times, the test time at the reference temperature was divided by the age conversion factor. For example, the 28 day equivalent age test for the hot batch should be tested at approximately 28 days/1.42, which equals approximately 20 days. The rest of the actual test times for the cold and hot mixtures are summarized in Table 4.2. Once the proper age was achieved, three of the nineteen cylinders were tested for compressive strength based on AASHTO T 22. The 76 author and his colleague performed all tests and were also ACI certified in ?Concrete Strength? testing during this project. Table 4.2 Compressive strength testing schedule Batch Identification Control Hot Cold 2ts 2ts 2ts 24 hr 18 hr 42 hr 48 hr 35 hr 84 hr 7 day 5 day 12 day 14 day 10 day 25 day Actual Age of Concrete at Testing 28 day 20 day 49 day Note: ts = final setting time Neoprene pads were used for all compressive strength tests. This system was chosen as it allowed the research team to test the strengths as soon as the specimens were removed from the curing tanks. Sulfur capping would have required the cylinders to be removed from the curing tanks two or more hours before testing. Pads were purchased from M.A. Industries, Inc. Pad durometer hardness was chosen based on ASTM C 1231, Table 1. Durometer 50, 60, or 70 pads were used depending on the expected level of compressive strength. A log of the number of uses of each set of pads was kept, and pads were discarded as required by ASTM C 1231. 77 All cylinders remained moist until time of testing. This was especially critical for the hot-cured cylinders at later ages. If they had not been kept moist, the heat would have caused the cylinders to dry quicker than a cylinder cured at cold or control temperatures. The potential moisture loss may seem small, but could have caused significant deviations in compressive strength test results if it had not been prevented. After the 28-day equivalent age test, the temperature and time data were downloaded from the i-Button. With the temperature and time history and the compressive strength test results, analysis with the maturity method could be performed. 4.4 SUMMARY In this chapter the experimental testing procedures were explained. Mixing procedures were given. For each batch, nineteen 6x12 in. cylinders were made, and the temperature history was recorded. The compressive strength testing schedule was given as the following six equivalent ages: 2 x final set time, 24 hr, 48 hr, 7 d, 14 d, and 28 d. The results of the testing described above are presented in the next chapter. 78 CHAPTER 5 PRESENTATION OF RESULTS In this chapter, the results are given for the tests described in Chapter 4 for all mixtures considered in this project. The fresh concrete properties are reported first, and then the hardened concrete properties are examined. The effects of the various temperature histories, supplementary cementing materials (SCMs), and other changes to mixture proportions are analyzed, when appropriate. 5.1 FRESH CONCRETE PROPERTY RESULTS The results of the fresh concrete testing are summarized in Table 5.1. There are some temperature effects on the air content and slump that could not be controlled because the chemical admixture dosages had to remain constant between batches of the same mixture. The target slump and air content are given in Chapter 4. Slump The Type I - 0.41 control batch had a slump of 1 in., which was less than the 2 to 4.5-in. target. This was deemed acceptable since the same mixture proportions were used for the 20% C, 30% C, 20% F, 30% F, 30% Slag, and 50% Slag mixtures, and additions of SCMs generally increase workability (Kosmatka, Kerkhoff, and Panarese 2002). The Class C fly ash and Class F fly ash mixtures did increase the slump, from 1 in. up to 6 in. for the control batches. However, the GGBF slag mixtures showed a small increase or no 79 change in slump relative to the Type I - 0.41 mixture. The Type I - 0.48 control batch had a slump of 6.25 in., but since the air content and density agreed with the other batches of that mixture, it was considered acceptable. Both of the prestressed concrete control mixtures, the Type III - 0.37 and 70/20/10 - 0.37, stayed under the 8-in. limit for slump. Air Content The Type I - 0.41 control batch was within the target air content range of 3 to 6%. The 20% F, 30% F, 30% Slag, and 50% Slag mixtures all showed a decrease in air content when compared to the Type I - 0.41 mixture. The Class C fly ash mixtures showed a slight increase in air content, but were considered acceptable once strength levels were checked. The cold batches from the Class C fly ash mixtures had high air contents and low densities, but this was uncontrollable as the mixture proportions were kept constant between batches of the same mixture. Also, the strength levels were comparable to the other batches of the same mixtures, which is reported in the next section. The air content for the control batches of the prestressed concrete mixtures were within the target air content of 3 to 6%. Also, note that the air contents of the mixtures with silica fume (70/20/10) were within the target range after using the procedure described in Section 4.1.2 to remove the excess entrapped air. Curing Temperature As soon as fresh concrete testing was finished and all cylinders were prepared, the cylinders were placed in their appropriate curing tanks. Due to heat from the hydration process, the concrete cylinders deviated from the curing tank?s cycle for the first few days. A graph of a typical cylinder temperature history versus the curing tank?s 80 temperature history is shown in Figure 5.1. From the figure, it can be seen that the research team placed the cold and hot batches in the curing tanks as the lime-water temperature in the tanks was rising, as explained in Section 4.2. The hot batch in this case reached the desired curing temperature range of 106?F to 90?F at approximately 36 hours, and the cold batch reached the desired curing temperature range of 55?F to 40?F at approximately 65 hours after placement. The importance of heating or cooling the raw materials to replicate ambient field temperatures can be seen in the figure. If all the mixtures had been batched at 73?F and then placed in their appropriate curing tanks, then the true effects of the curing temperatures on the strength gain behavior of the mixtures could not be evaluated. 81 Table 5.1: Fresh concrete properties of all mixtures Mixture ID Batch ID Slump (in.) Temperature (?F) Air Content (%) Density (lb/ft3) 4.50 148 6.00 145 Type I - 0.41 Hot Control Cold 2.50 1.00 2.75 96 68 61 5.75 142 3.75 150 7.50 146 Type I - 0.44 Hot Control Cold 1.50 4.00 3.25 102 73 58 9.00 146 4.50 147 4.75 147 Type I - 0.48 Hot Control Cold 2.75 6.25 3.00 104 74 56 5.25 146 2.25 149 2.75 150 20% F Hot Control Cold 2.50 5.75 6.50 100 70 53 3.00 147 2.50 149 2.00 149 30% F Hot Control Cold 1.25 3.50 4.00 104 70 53 2.75 148 4.00 147 6.75 144 20% C Hot Control Cold 2.50 6.00 7.25 101 71 55 8.00 140 3.50 147 6.25 144 30% C Hot Control Cold 2.75 6.00 7.50 101 74 56 8.00 141 3.75 150 4.50 149 30% Slag Hot Control Cold 0.50 1.50 2.00 100 74 53 7.75 146 3.00 149 3.50 149 50% Slag Hot Control Cold 1.00 1.00 3.00 98 72 50 4.50 147 3.75 151 3.50 150 Type III - 0.37 Hot Control Cold 1.00 5.25 8.75 101 71 55 9.50 145 2.75 149 5.50 146 Type III - 0.44 Hot Control Cold 1.25 3.50 5.75 98 76 55 8.25 144 3.25 148 3.00 148 70/20/10 - 0.37 Hot Control Cold 2.00 7.75 8.75 98 71 56 3.50 148 2.00 147 2.50 148 70/20/10 - 0.44 Hot Control Cold 2.50 3.75 5.75 105 75 51 3.25 146 82 30 40 50 60 70 80 90 100 110 120 0 24 48 72 96 120 Concrete Age (hours) Temperature (?F) Figure 5.1: Typical temperature history of concrete cylinders 5.2 HARDENED CONCRETE TEST RESULTS The results from the compressive strength tests for all mixtures can be found in Appendix A. The average test result for each age was plotted versus chronological time for all three batches of all thirteen mixtures in Figures 5.2 through 5.14. A least-squares regression analysis was performed to fit the best hyperbolic strength-age function to the strength versus age data as prescribed by ASTM C 1074. The equation is as follows: )(1 )( o o u ttk ttkSS ?+ ?= Previously Equation 2.8 where, S = average compressive strength at age, t, (psi), Cylinder Curing Tank 83 Su = limiting strength (psi), t = test age (hr), to = age when strength deve lopment is assumed to begin (hr), and k = the rate constant (1/hour). The actual test data were compared to estimated strength values computed from the hyperbolic strength-age equation, and the difference between the actual and estimated strength was computed. The differences were then squared, to eliminate negative values, and summed. This was performed for all three batch temperatures of all mixtures. The ?Solver? function in Microsoft Excel was utilized to minimize the sum of the squares of the errors by varying the k, Su, and to values. The least-squares k, Su, and to values determined for all mixtures are reported in Table 5.2. The best-fit to value was computed as negative for some batches, such as the Type I - 0.44 hot and 20% F hot and control batches, due to extremely rapid initial strength gain. These were corrected because they are not realistic values. As it is defined above, to is the time at which the concrete begins to gain strength. Thus, to should never be less than zero, and whene ver this error occurred, to was limited to 0 hrs and the least- squares regression was re-solved. After examination the strength versus age plots in Figures 5.2 through 5.14, it is obvious that all of the mixtures were affected by the various temperature histories. If this was not true, then all the test points for one mixture would fall on a single curve, and there would be no need to apply the maturity method to estimate the strength. As a 84 precursor to estimating the strengths using the maturity method, it is important to evaluate and quantify the effect of the temperature histories on the strength-age behavior. For this, the amount of crossover, as described in Section 2.3.2, and the amount of temperature correction required by the different mixtures will be examined. 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.2: Compressive strength versus age results for Type I - 0.41 mixture 85 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.3: Compressive strength versus age results for Type I - 0.44 mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.4: Compressive strength versus age results for Type I - 0.48 mixture 86 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.5: Compressive strength versus age results for 20% F mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.6: Compressive strength versus age results for 30% F mixture 87 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.7: Compressive strength versus age results for 20% C mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.8: Compressive strength versus age results for 30% C mixture 88 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.9: Compressive strength versus age results for 30% Slag mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.10: Compressive strength versus age results for 50% Slag mixture 89 0 2,000 4,000 6,000 8,000 10,000 12,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold data Hot Data Figure 5.11: Compressive strength versus age results for Type III - 0.37 mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.12: Compressive strength versus age results for Type III - 0.44 mixture 90 0 2,000 4,000 6,000 8,000 10,000 12,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.13: Compressive strength versus age results for 70/20/10 - 0.37 mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data Figure 5.14: Compressive strength versus age results for 70/20/10 - 0.44 mixture 91 Table 5.2: Regression values for strength-age curves Mixture ID Batch ID to (hr) Su (psi) k (hr-1) 0.082 0.047 Type I - 0.41 Hot Control Cold 1.4 8.3 15.3 5,710 6,060 6,530 0.026 0.041 0.018 Type I - 0.44 Hot Control Cold 0.0 1.4 11.9 5,530 5,950 6,590 0.011 0.034 0.015 Type I - 0.48 Hot Control Cold 0.9 0.8 8.3 5,260 5,970 7,870 0.006 0.041 0.019 20% F Hot Control Cold 0.0 0.0 11.2 7,020 7,100 6,840 0.015 0.028 0.015 30% F Hot Control Cold 0.0 0.0 11.9 7,150 6,840 6,550 0.012 0.054 0.026 20% C Hot Control Cold 1.8 5.5 14.2 6,420 6,770 6,820 0.013 0.039 0.017 30% C Hot Control Cold 5.7 7.0 14.1 7,350 7,230 7,480 0.007 0.060 0.020 30% Slag Hot Control Cold 3.4 0.4 4.5 6,240 7,550 8,390 0.007 0.031 0.012 50% Slag Hot Control Cold 4.0 4.5 1.8 7,290 8,850 10,740 0.003 0.142 0.063 Type III - 0.37 Hot Control Cold 3.6 3.3 11.4 8,860 10,730 10,690 0.027 0.081 0.044 Type III - 0.44 Hot Control Cold 0.0 6.5 14.4 5,550 6,180 7,260 0.015 0.037 0.015 70/20/10 - 0.37 Hot Control Cold 0.2 0.0 0.0 9,080 11,770 11,530 0.007 0.027 0.009 70/20/10 - 0.44 Hot Control Cold 0.0 0.0 2.1 5,670 8,220 8,510 0.004 92 To quantify crossover, the late-age strength difference between the control batch and the hot and cold batches, as well as the amount of time until crossover, was examined. The amount of crossover will affect how accurate the strength estimations using the maturity method analyzed in Chapter 6 can be. Also, the amount of temperature correction required for convergence of the three individual strength-age curves onto one unique strength-maturity curve will be discussed. This affects the selection of datum temperatures or activation energies, as discussed in Section 2.4, which are needed in to apply the maturity method to estimate strengths as discussed in Chapter 6. The amount of crossover was quantified from the best-fit limiting strength (Su) values reported in Table 5.2, from the following equation: Crossover Factor (COF) 100)( ??= controlu controluu S SSCOF Equation 5.1 where, COF = Crossover factor, %, Su = Limiting strength found from regression analysis for hot or cold batch (psi), and Su control = Limiting strength found from regression analysis for control batch (psi). The crossover factor (COF) gives the percent difference between the control batch?s limiting strength and the hot or cold batch?s limiting strength. The COF is negative if 93 there was crossover for the hot batches and positive if there was crossover for the cold batches. Otherwise, no crossover occurred. The crossover factors for the hot and cold batches of all mixtures are given in Table 5.3. If no crossover occurred, the term ?N/A? is used as ?not applicable.? Also in the table are the times at which crossover occurred, in hours and days. The mixtures are compared and analyzed based on the crossover data and temperature sensitivity and the effect of SCMs, varying cement types, and water-to- cementitious materials ratios. 94 Table 5.3: Crossover factors and crossover time for all mixtures Mixture ID Batch ID COF (%) Crossover Time, hr (day) Hot -6 160 (7) Type I - 0.41 Cold 8 250 (10) Hot -7 390 (16) Type I - 0.44 Cold 11 290 (12) Hot -12 260 (11) Type I - 0.48 Cold 32 230 (10) Hot -1 2300 (96) 20% F Cold N/A N/A Hot N/A N/A 30% F Cold N/A N/A Hot -5 360 (15) 20% C Cold 1 5300 (220) Hot N/A N/A 30% C Cold 3 2300 (96) Hot -17 140 (6) 30% Slag Cold 11 790 (33) Hot -18 230 (10) 50% Slag Cold 21 1100 (46) Hot -17 40 (2) Type III - 0.37 Cold N/A N/A Hot -10 100 (4) Type III - 0.44 Cold 18 250 (10) Hot -23 110 (5) 70/20/10 - 0.37 Cold N/A N/A Hot -31 130 (5) 70/20/10 - 0.44 Cold 3 3400 (140) Note: N/A = No crossover occurred 95 5.2.1 Crossover Effect Straight Type I Mixtures From Figures 5.2 through 5.4 and Table 5.3, it can be seen that all straight Type I cement mixtures experienced crossover. The Type I - 0.41 and Type I - 0.44 hot batches both had long-term strength losses of about 7% relative to their respective control batches. The Type I - 0.41 mixture had crossover for the hot batch at approximately 7 days, while the Type I - 0.44 hot batch had crossover much later, at 16 days. The Type I - 0.48 mixture had the highest long-term strength loss due to high curing temperatures, for all the straight Type I cement mixtures at 12%, and crossover occurred at approximately 11 days. The crossover factors for the cold mixtures are not as much of a concern as for the hot batches, since the maturity method would estimate a lower strength value than that of the actual concrete and thus would be conservative. Effect of Replacement Dosages of SCMs on the Type I - 0.41 mixture The effect of replacement dosages of SCMs on the Type I - 0.41 mixture can be seen for the 20% F, 30% F, 20% C, 30% C, 30% Slag, and 50% Slag mixtures in Table 5.3 and Figures 5.5 through 5.10. From the table and figures, it can be seen that the replacement of cement with Class F fly ash at doses of 20% and 30% effectively eliminated the crossover effect. (The 20% F hot batch had a 1% long-term strength loss due to crossover that only occurred after 96 days; thus for all practical purposes, the crossover effect was eliminated.) The replacement of cement with Class C fly ash at 20% by mass did not completely eliminate the crossover effect for the hot batch, when compared to the Type I - 0.41 mixture, but it did delay the time until crossover from 7 days to 15 days. The replacement dosage of 30% Class C fly ash by mass did effectively 96 eliminate the crossover effect when compared to the Type I - 0.41 mixture. The replacement doses of GGBF slag increased the long-term strength loss due to high curing temperatures relative to the loss for the Type I - 0.41 mixture. The crossover factor went from 6% strength loss, for the Type I - 0.41 mixture, to 17% strength loss for the 30% Slag hot batch and 18% for the 50% Slag hot batch. These long-term strength losses are important to note because the maturity method cannot correct for them, and the accuracy of estimating strengths using the maturity method will suffer accordingly. Effect of Replacement Dosages of SCMs and Varying Cement Types on the Type I - 0.44 mixture The effect of changing cement types and replacement dosages of SCMs on the Type I - 0.44 mixture (ALDOT standard mixture A1-c) can be seen from the Type III - 0.44 and 70/20/10 - 0.44 mixtures in Figures 5.12 and 5.14 and Table 5.3. The effect of changing from Type I cement to Type III cement increased the long-term strength loss attributable to high curing temperatures slightly, from 7% to 10%, but also decreased the time until crossover occurred, from 16 days to 4 days. Replacing the Type I cement with 20% Class F fly ash and 10% silica fume greatly increased the long-term strength loss due to high curing temperatures from 7% to 31% strength loss when compared to their respective control batches. The time until crossover also decreased from 16 days for the Type I - 0.44 to 5 days for the 70/20/10 - 0.44 mixture. Prestressed Concrete Mixtures The prestressed concrete mixtures (Type III and the ternary mixture with water- to-cementitous materials ratios of 0.37) had some of the worst long-term strength loss effects due to high curing temperatures and earliest crossover ages. This can be seen in 97 Figures 5.11 and 5.13 and Table 5.3. The Type III - 0.37 hot batch had a long-term strength loss of 17% when compared to its respective control batch. Also, the plateau of strength of the hot batch, seen in Figure 5.11, is quickly reached at approximately 40 hrs. Thus, any strength estimated by the maturity method, for this mixture, after 40 hrs will probably be overestimated (unconservative). The ternary-blend prestressed concrete mixture had a long-term strength loss attributable to high curing temperatures of 23% relative to its respective control batch. Also, this strength loss began at approximately 5 days. 5.2.2 Temperature Sensitivity The amount of correction due to temperature effects needed for the maturity method refers to how far the hot and cold batches must shift on the time scale (once converted to maturity) to converge onto an unique strength-maturity relationship. Some common temperature sensitivity trends are shown schematically in Figure 5.15. Some mixtures may have a high temperature sensitivity, which requires a large correction by the maturity method, as shown by the first schematic in the figure. Others may require a small correction, which would be referred to as low temperature sensitivity. There could also be some mixtures that have different temperature sensitivities for the hot and cold batches, i.e. the hot batch could require a smaller temperature correction than the cold batch, which is shown in the final schematic. The temperature sensitivities for the hot and cold batches of each mixture were evaluated qualitatively based on visual observations from Figures 5.2 through 5.14. If the crossover effect occured, the temperature sensitivity is only discussed for ages before the 98 point of crossover. In Chapter 6, these temperature sensitivity trends are referred to when values of datum temperatures and activation energies are determined. Figure 5.15: Schematic of various degrees of temperature sensitivity Straight Type I Mixtures From Figure 5.2, it appears that the Type I - 0.41 mixture does not need much of a correction due to temperature effects, since all three strength-age curves are relatively close to each other. For the Type I - 0.44 mixture the hot batch may require a slightly higher temperature sensitivity value than the cold batch, but overall the mixture does not require a high temperature sensitivity value. The Type I - 0.48 mixture also only requires a small temperature correction. Effect of Replacement Dosages of SCMs on Type I - 0.41 Mixture The replacement of cement with 20% Class F fly ash increased the temperature sensitivity of the Type I - 0.41 mixture slightly. Increasing the replacement dosage to ?High? Cold Temperature Sensitivity Concrete Age Concrete Age Concrete Age Hot Batch Control Batch Cold Batch ?High? Temperature Sensitivity ?Low? Temperature Sensitivity ?Low? Hot Temperature Sensitivity ?Mixed? Temperature Sensitivity Compressive Strength 99 30% increased the temperature correction needed for the hot batch, but not for the cold batch. The replacement of cement with 20% Class C fly ash increases the temperature sensitivity of the cold batch significantly, but not the hot batch. However, increasing the dosage to 30% Class C fly ash further increased the temperature sensitivity of both batches substantially. The replacement of cement with 30% GGBF slag greatly increases the temperature correction required for the cold batch and does not affect the hot batch?s sensitivity. Increasing the dosage of GGBF slag to 50% only increases the temperature correction needed for the cold batch and again does not affect the hot batch?s temperature sensitivity significantly. Effect of Replacement Dosages of SCMs and Varying Cement Types on Type I - 0.44 Mixture For the Type I - 0.44 mixture, changing the Type I cement to Type III cement decreases the temperature correction needed for the hot batch and increases the correction needed for the cold batch. Replacing the Type I cement with 20% Classs F fly ash and 10% silica fume decreases the temperature sensitivity of the hot batch and significantly increases the temperature sensitivity of the cold batch. Prestressed Concrete Mixtures The temperature correction required for the Type III - 0.37 hot batch is minimal. The cold batch requires a moderate correction value. The 70/20/10 - 0.37 hot mixture requires a low temperature correction value, while the cold batch requires a larger correction. 100 5.3 SUMMARY In this chapter, the results for testing of all thirteen mixtures considered in this project were discussed. The fresh concrete properties were reported, and though there were some deviations in slump and air content from the desired values due to temperature and the replacement dosages of SCMs, the concrete was of good quality. The compressive strength versus concrete age graphs were given and compared. Hyperbolic strength-age regression curvess were fit to the data, and the best-fit parameters were reported. It was obvious that all mixtures were affected in some way by the various curing temperature histories. The crossover factor was defined and computed for all mixtures. All mixtures were then analyzed based on the amount of crossover and time until crossover occurred. Based on the work documented in this chapter, the following conclusions can be made: ? All straight cement mixtures had crossover, with long-term strength losses ranging from 7% to 12%, and crossover occurring between 7 days and 16 days. ? The replacement of cement with 20 and 30% Class F fly ash for the Type I - 0.41 mixture effectively eliminated the crossover effect. ? The replacement of cement with 20% Class C fly ash for the Type I - 0.41 mixture delayed, but did not completely eliminate, the crossover effect. An increased replacement dosage of 30% Class C fly ash for the Type I - 0.41 mixture effectively eliminated crossover. 101 ? The replacement of cement with 30% or 50% GGBF slag for the Type I - 0.41 mixture increased the long-term strength loss, or crossover, attributed to high curing temperatures from 6% to more than 17% in some cases. ? Changing the cement from Type I to Type III for the Type I - 0.44 mixture increased the crossover effect only slightly, but decreased the time at which crossover occurs from 16 days to 4 days. ? The replacement of cement with 20% Class F fly ash and 10% silica fume for the Type I - 0.44 mixture increased the long-term strength loss attributed to high curing temperatures from 7% to 31% and decreased the time at which crossover occurred. ? The Type III - 0.37 mixture had a 17% strength loss due to the crossover effect and crossover occurred less than two days after mixing for the hot batch. ? The ternary blend prestressed concrete mixture had a long-term strength loss of 23% for the hot batch. Finally, the mixtures were examined based on the visible amount of temperature correction, or temperature sensitivity, required to cause convergence of the hot and cold batches onto one strength-maturity relationship. A summary of these conclusions follows: ? The Type I - 0.41 and Type I - 0.48 mixtures have low temperature sensitivity for both hot and cold batches. 102 ? The Type I - 0.44 mixture had low temperature sensitivity, possibly with a slightly higher sensitivity for the hot batch. ? The replacement of cement with 20% Class F fly ash for the Type I - 0.41 mixture increased the temperature sensitivity slightly and increasing the dose to 30% Class F fly ash increased the hot batch?s temperature sensitivity. ? The replacement of cement with 20% Class C fly ash for the Type I - 0.41 mixture increased the temperature sensitivity of the cold batch, and increasing the dose to 30% Class C fly ash increased both batches? temperature sensitivities significantly relative to that of the Type I - 0.41 sensitivities. ? The replacement of cement with 30 and 50% GGBF slag for the Type I - 0.41 mixture increased the temperature sensitivity of the cold batch, but did not affect the hot batches. ? Changing the cement from Type I to Type III for the Type I - 0.44 mixture increased the temperature sensitivity for the cold batch and decreased that for the hot batch. ? The replacement of cement with 20% Class F fly ash and 10% silica fume for the Type I - 0.44 mixture increased the temperature sensitivity for the cold batch and decreased that for the hot batch. ? The Type III - 0.37 mixture?s hot batch had little to no temperature sensitivity, while the cold batch required a higher temperature sensitivity. ? The ternary blend prestressed concrete mixture (70/20/10 - 0.37) had a low temperature sensitivity for the hot batch and a higher value for the cold batch. 103 Now that the effects of the various curing temperature histories, types and doses of SCMs, cement types, and water-to-cementitious material ratios have been established, the accuracy of the maturity method to estimate concrete strength is evaluated in the next chapter. 104 CHAPTER 6 ANALYSIS OF RESULTS The effect of various curing temperatures on the development of compressive strength was discussed in Chapter 5. Next, the accuracy of the maturity method to estimate the strength development of hardening concrete is evaluated. As stated in the Chapter 2, there are two well known functions used to compute the maturity index. These are known as the Nurse-Saul maturity (NSM) and Arrhenius maturity (AM) functions. In this chapter, the accuracy of these two maturity functions to estimate concrete strength is examined. Different methods are proposed in order to find the simplest and most accurate way to compute the maturity index and accurately estimate the concrete strength. As described in Chapter 2, a strength-maturity relationship is mixture-specific. This is due in part to the fact that variables such as water-to-cementitious materials ratio, SCM type and dosage, cement type, etc. all affect the strength and temperature development in a complex manner. The mixture?s temperature sensitivity refers to the degree to which the concrete will be affected by curing temperatures different than the reference temperature and how much correction will be needed to convert the strength- age curves into one unique strength-maturity relationship. This was previously described with Figure 5.15. In Section 5.2.2 it was shown that some mixtures such as the Type I - 0.41 had a low temperature sensitivity, while others like the 30% C mixture needed a 105 much larger temperature correction in order to converge the strength-age curves to a single strength-maturity relationship. Other mixtures such as the GGBF slag mixtures had a low temperature sensitivity for the hot batch but a very high sensitivity for the cold batch. The mechanisms that account for temperature sensitivity when using the maturity method are the datum temperature and activation energy, depending on the maturity function used. These values were defined as temperature sensitivity values in Section 2.4.2. In this chapter, several methods are presented to find or compute proper temperature sensitivity values that maximize the accuracy of the maturity method to estimate concrete strength. Based on these different methods, the most accurate and simplest method for computing maturity and estimating concrete strength is recommended. The strength data were ana lyzed using the current ASTM C 1074 (2004) specification, which explains how to find the mixture-specific activation energies or datum temperatures for any mixture. The data were then analyzed using the same methods as the ASTM C 1074 specification, with some modifications. After the results of the analyses are presented, results will be presented based on analysis ofa simplified approach using constant temperature sensitivity parameters for all the mixtures. Finally, an evaluation of formulaic determination of temperature sensitivities will be presented. In each case, three mixtures will be shown as examples, and an assessment of the accuracy of the maturity method to estimate the compressive strength for all 13 mixtures will be discussed. 106 6.1 ANALYSIS OF DATA BASED ON THE ASTM C 1074 METHOD ASTM C 1074 (2004) has several different methods to determine a mixture- specific datum temperature or activation energy. This is described in detail in ASTM C 1074 under Annex A1. In this study, the option Section A1.1.8 was used, which is considered the most accurate and requires the use of regression analysis techniques. This method is identical to the method shown in Section 2.4.1, which uses the rate constants from the strength-age data to determine temperature sensitivity values. The k, Su, and to values from the hyperbolic strength-age formula, already computed in Chapter 5 and summarized in Table 5.2, are utilized for this method. 6.1.1 Nurse-Saul Maturity Function As described in Chapter 2, the equation for computing the maturity index using the Nurse-Saul maturity (NSM) function is: tTTM t oc ???= ? 0 )( Previously Equation 2.1 where, M = maturity index at age t, (Temperature-Time Factor), (?F?hours), Tc = Average concrete temperature during the time interval (?F), Dt = a time interval (hours), and To = datum temperature (?F). 107 In this equation, all values are known except To, the datum temperature. As explained before, this value is related to the temperature sensitivity of a mixture and will affect how well the maturity method can estimate the strength at all curing temperatures. In this section, the mixture-specific datum temperature, based on ASTM C 1074 (2004) methods, for all mixtures is given. Later, in Section 6.1.3, an analysis of the accuracy of using these mixture-specific datum temperatures with the NSM function will be given. In order to find the mixture-specific datum temperatures, the k-values from the hyperbolic strength-age function for each batch of one mixture are plotted versus temperature. As explained by Carino (1991) and ASTM C 1074 (2004), the mixture- specific datum temperature is the x-intercept of a linear trendline fitted through the points. An example of this process is shown in Figure 6.1. This procedure thus gives the temperature that corresponds to a zero rate of strength development. From the plot, the three k-values can be seen for each batch temperature for the Type I - 0.41 mixture. The cold batch had a k-value of approximately 0.025 hr -1, the control batch?s k-value was approximately 0.046 hr -1, and the hot batch?s k-value was about 0.082 hr -1. A linear trendline was fit to the data, and, after extending the line to the x-axis, the mixture- specific datum temperature for the Type I - 0.41 mixture based on ASTM C 1074 methods was calculated to be 23?F (-5?C). This procedure was used on all thirteen mixtures, and the resulting mixture- specific datum temperatures are presented in Table 6.1. The values with asterisks represent mixture-specific datum temperatures that are greater than some of the temperatures reached during the cold curing conditions. Recall that in the Nurse-Saul maturity function the temperature of the concrete is subtracted from the datum 108 temperature. However, if the datum temperature is lower than the curing temperature, the maturity index becomes negative. As Saul (1951) stated, the datum temperature is the temperature at which concrete ceases to gain strength. However, neither maturity nor strength should ever be subtracted when temperatures fall below the datum temperature. For this reason, it was assumed that the cold mixtures that had curing temperatures less than the datum temperatures accumulated zero maturity during the time intervals where the concrete temperature was less than the datum temperature. y = 0.0011x - 0.0246 R2 = 0.9991 0.00 0.02 0.04 0.06 0.08 0.10 20 40 60 80 100 Average Curing Temperature (?F) Rate Constant (1/hr) Figure 6.1: Method used to find the best-fit datum temperature, ASTM C 1074 method for Type I ? 0.41 mixture After each mixture-specific datum temperature was found, the maturity index could be computed using the NSM function for all mixtures and batches. To illustrate To = 23?F 109 how the maturity method works with different cementitious systems and mixtures, three examples will be used throughout this chapter: 1. The 20% F mixture, a mixture that has no long-term strength loss effects due to curing at high temperatures, 2. The Type I - 0.41 mixture, a mixture that has some long-term strength loss effects due to curing at high temperatures, and 3. The Type III - 0.37 mixture, a mixture that has severe long-term strength loss effects due to curing at high temperatures. Table 6.1: Mixture-specific datum temperatures based on ASTM C 1074 Mixture ID Datum Temperature, To (?F (?C)) Type I - 0.41 23 (-5) Type I - 0.44 35 (2) Type I - 0.48 39 (4) 20% F 23 (-5) 30% F 19 (-7) 20% C 37 (3) 30% C 42 (5)* 30% Slag 46 (8)* 50% Slag 47 (9)* Type III - 0.37 38 (3) Type III - 0.44 37 (3) 70/20/10 - 0.37 39 (4) 70/20/10 - 0.44 43 (6)* * Value higher than some temperatures reached during cold temperature history For ease of comparison, the compressive strength versus age plots, shown in Chapter 5, are repeated above the accompanying compressive strength versus maturity 110 plots. These plots can be seen in Figures 6.2 through 6.4. Strength-maturity (S-M) plots for all other mixtures can be seen in Appendix B. The strength-maturity (S-M) relationships shown in the figures were calculated using the hyperbolic S-M equation, discussed in Chapter 2, fit by least-squares regression to the control-batch data. The equation is repeated here: )(1 )( o o u MMk MMkSS ?+ ?= Previously Equation 2.5 where, S = average compressive strength at maturity, M, (psi), M = maturity index (?F?hr or hr), Su = limiting compressive strength (psi), Mo = maturity index when strength development is assumed to begin (?F?hr or hr), and k = the rate constant [1/(?F?hr or hr)]. This equation is also used later to determine the accuracy of the strength estimated by the maturity method. From Figures 6.2 through 6.4, it can be seen that mixtures that have little to no long-term strength loss due to curing at high temperatures, such as the 20% F mixture, have the least deviation from their S-M relationship. This means the maturity method will produce low errors when used to estimate strengths for these mixtures. From Figure 6.3 it can be seen that for a mixture that has a relatively small crossover effect, the maturity method will estimate strength well until the point of crossover, i.e. approximately 160 to 250 hours of actual age, in this case. After crossover occurs the 111 hot strengths are overestimated (un-conservative) and the cold strengths are underestimated (conservative). And finally, for a mixture with a severe crossover effect, as in Figure 6.4, the maturity method will estimate strength well until crossover and then grossly overestimate strengths for hot mixtures, while in this case the cold mixture?s strengths fit the strength-maturity curve very well. This illustrates the fact that the maturity method can only shift data values on the age (horizontal) scale and cannot correct for long-term strength loss or gain due to the crossover effect. The observations made in this paragraph are also valid for all the other mixtures evaluated. 112 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 10,000 20,000 30,000 40,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure 6.2: 20% F mixture, (a) compressive strength versus age, (b) strength-maturity plot, ASTM C 1074 method To = 23?F 113 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure 6.3: Type I - 0.41 mixture, (a) compressive strength versus age, (b) strength- maturity plot, ASTM C 1074 method To = 23?F 114 0 2,000 4,000 6,000 8,000 10,000 12,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold data Hot Data (a) 0 2,000 4,000 6,000 8,000 10,000 12,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure 6.4: Type III - 0.37 mixture, (a) compressive strength versus age, (b) strength- maturity plot, ASTM C 1074 method To = 38?F 115 Comparison of Results to Previously Published Findings When the datum temperatures found in this study are compared to previously published reports and literature, some similarities can be found as well as some differences. Table 6.2 repeats the previous Table 2.1. As Carino and Tank (1992) found in their study, none of the computed To values were equal to the commonly used 14?F (- 10?C), which was also true in this study. The lowest values reported in Table 6.1, 23?F (-5?C) and 19?F (-7?C), were calculated for the Type I - 0.41 mixture and the 20 and 30% Class F fly ash mixtures. Carino and Tank also found 23?F (-5?C) for their Type I + 20% fly ash mixture (class of fly ash was not specified). Also, the mixtures with GGBF Slag dosages in both studies required higher datum temperatures that ranged between 46?F (8?C) and 50?F (10?C). However, the datum temperature of 37?F (3?C) found in this study for the Type III mixtures is lower than Carino and Tank?s value of 45?F (7?C). Also in Carino and Tank?s study, as the water-to-cement ratio increased for the straight cement mixtures, it was found that the datum temperatures generally decreased. In this study, as the water-to-cement ratio increased for the straight Type I cement mixtures, so did the datum temperatures. This could be related to the diminishing need of water- reducer required as the water-to-cement ratio increased (See Table 3.5). Alternatively, the fact that fluctuating curing conditions were used may have produced differences in behavior, since Carino and Tank used isothermal curing conditions. 116 Table 6.2: Datum temperature values proposed by Carino and Tank (1992) Datum Temperature, To ( F ( C)) Cement Type w/c = 0.45 w/c = 0.60 Type I Type II Type III Type I + 20% Fly Ash Type I + 50% Slag Type I + Accelerator Type I + Retarder 52 (11) 48 (9) 45 (7) 23 (-5) 46 (8) 46 (8) 41 (5) 48 (9) 43 (6) 45 (7) 32 (0) 50 (10) 48 (9) 41 (5) 6.1.2 Arrhenius Maturity Function Next, the ASTM C 1074 (2004) method for determining mixture-specific temperature sensitivity values was used for the Arrhenius maturity (AM) function. As described in Section 2.1, the Arrhenius-based equation for converting chronological time to equivalent age is defined by the following equation: tet t TTRE e rc ??= ? ?? ? ?? ? +?+ ? 0 273 1 273 1 Previously Equation 2.2 where, te = equivalent age at the reference curing temperature (hours), Tc = average concrete temperature during the time interval, Dt, (?C), Tr = reference temperature, 23?C, E = activation energy, J/mol, and R = universal gas constant, 8.314 J/(mol?K). 117 Again, all of the values are known except for the temperature sensitivity term, which is the activation energy, E, in this method. The k-values from the strength-age functions were again used. As explained in Section 2.4.1, in order to find the mixture-specific activation energy, the natural logarithms of the k-values were plotted versus the inverse of the absolute batch temperatures, in Kelvin units. Figure 6.5 shows an example of this plot, often called the Arrhenius plot, for the Type I - 0.41 mixture. To find the mixture- specific activation energy, the slope of the fitted linear trendline was determined. The mixture-specific activation energy is the negative of the slope of the trendline multiplied by the universal gas constant, 8.314 J/(K?mol). As shown in Figure 6.5, the mixture- specific activation energy for the Type I - 0.41 mixture was calculated to be 28,600 J/mol. Note that all the activation energy values were rounded to the nearest 100 J/mol. This procedure was used on all thirteen mixtures, and the resulting mixture-specific activation energies are summarized in Table 6.3. The same three mixtures that were shown in Section 6.1.1 above will be analyzed in this section using the AM function. Their strength-maturity (S-M) plots are found in Figures 6.6, 6.7, and 6.8. The remaining mixtures? S-M plots can be found in Appendix C. From Figures 6.6 to 6.8, it can be seen that the trends are similar to those found using the NSM function; that is, mixtures with little to no crossover were modeled well with these S-M relationships, and mixtures with more crossover generally were accurately modeled by the S-M relationships only until the point of crossover. Again, maturity can only translate values on the horizontal axis of the strength versus maturity plots. Later in this section, assessment of the accuracy of the two maturity functions using the ASTM C 1074 methods is discussed. 118 y = -3442.61x + 8.64 R2 = 0.97 -3.8 -3.6 -3.4 -3.2 -3.0 -2.8 -2.6 -2.4 0.0032 0.0033 0.0034 0.0035 0.0036 1/Temperature (1/K) Ln(Rate Constant) [Ln(1/hr)] Figure 6.5: Method for computing activation energy based on ASTM C 1074 for the Type I - 0.41 mixture Table 6.3: Mixture-specific activation energies based on the ASTM C 1074 method Mixture ID Activation Energy, E (J/mol) Type I - 0.41 28,600 Type I - 0.44 34,800 Type I - 0.48 42,300 20% F 25,800 30% F 23,200 20% C 37,800 30% C 45,100 30% Slag 55,700 50% Slag 61,500 Type III - 0.37 41,400 Type III - 0.44 42,900 70/20/10 - 0.37 42,600 70/20/10 - 0.44 47,300 E = 3442 K [8.314 J/(K?mol)] E = 28,600 J/mol 119 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.6: Strength-maturity plot, ASTM C 1074 method for 20% F mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 200 400 600 800 1,000 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.7: Strength-maturity plot, ASTM C 1074 method for Type I - 0.41 mixture E = 28,600 J/mol E = 25,800 J/mol 120 0 2,000 4,000 6,000 8,000 10,000 12,000 0 200 400 600 800 1,000 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.8: Strength-maturity plot, ASTM C 1074 method for Type III - 0.37 mixture Comparison of Results to Previously Published Findings Next the values obtained from the ASTM C 1074 mixture-specific activation energy analysis will be compared to other published values for similar mixtures. Again, for ease of comparison, the values reported from Carino and Tank?s study in 1992 are summarized in Table 6.4. As with the datum temperature values, the Type I - 0.41 mixture and the 20 and 30% Class F fly ash mixtures had the lowest temperature sensitivity values of 28,600 J/mol to 23,200 J/mol in this study. Carino and Tank also obtained values around 30,000 J/mol for their Type I + 20% Fly Ash mixture. However, their value for a straight Type I cement is much higher than the value of 29,000 J/mol found in this study. As reported in Section 2.4.2, the following activation energy values have also been recommended previously for use with straight Type I cement mixtures: E = 41,400 J/mol 121 ? 40,000 to 45,000 J/mol for a Type I cement mixture with no admixtures or additions (ASTM C 1074 2004). ? 42,000 to 44,000 for a Type I cement mortar or concrete mixture (Carino 1991). Table 6.4: Activation energy values proposed by Carino and Tank (1992) Activation Energy, E (J/mol) Cement Type w/c ratio = 0.45 w/c ratio = 0.60 Type I Type II Type III Type I + 20% Fly Ash Type I + 50% Slag Type I + Accelerator Type I + Retarder 64,000 51,000 44,000 30,000 45,000 45,000 39,000 48,000 43,000 44,000 31,000 56,000 50,000 39,000 These values are also higher than some of the activation energies found for straight Type I cement mixtures in this study. However, the Type I - 0.48 mixture had an activation energy of 42,300 J/mol, which does agree with past research. This difference could be due to the removal of the water-reducing admixture as the water-to-cement ratio increased, or to the use of fluctuating curing conditions. As was the case with the datum temperature values, the activation energies increased as the water-to-cement ratio increased for the straight Type I cement mixtures. The values for the GGBF Slag mixtures have higher values than those reported by Carino and Tank at 45,000 J/mol for the lower water-to-cement ratio Type I/II + 50% Slag mixture. However, the 55,000 J/mol to 61,500 J/mol found in this study is closer to the values recommended by Carino (1991) of 49,000 J/mol and RILEM (Springenschmid 1998) of 50,000 J/mol. 122 6.1.3 Accuracy of Maturity Functions Based on the ASTM C 1074 Method Finally, the accuracy of using the ASTM C 1074 method to determine mixture- specific temperature sensitivity values was analyzed and is reported here. Reported errors are based on the difference between the actual compressive strength test values and the strength estimated from the hyperbolic strength-maturity relationship (Equation 2.5) using the NSM or AM functions. Average absolute errors (which have been used by Tank and Carino [1991]) were calculated as well from the following formula: Average Absolute Error (AAE) n StrengthStrengthEst AAE i n i i? ? = . Equation 6.1 where, AAE = Average absolute error for n strength estimations (psi), Strengthi = Compressive strength test result, for the ith maturity index, Mi, (psi), Est. Strengthi = Estimated strength from hyperbolic strength- maturity function, at the same maturity, Mi, (psi), and, n = number of strength estimations (6 for each batch; 18 for the total AAE for each mixture). The errors, based on the NSM and AM functions, for the three mixtures considered in this chapter, can be found in Tables 6.5 to 6.7 and 6.8 to 6.10, respectively. The 123 remaining error tables can be found in Appendix D. The AAEs for each batch are reported as well as the percent error of the strength estimate, which was computed from the following equation: Error of Estimate (EoE) 100. ??= i ii Strength StrengthStrengthEstEoE Equation 6.3 where, EoE = Error of estimate (%). The error of estimate (EoE) gives the percent of over- or underestimation produced by the S-M functions. A positive EoE indicates that the estimated strength was more than the measured strength, which means the strength was overestimated. A negative EoE indicates that the estimated strength was less than the measured strength, which means the strength was underestimated. To better visualize this, after each error table, the estimated strengths are plotted versus the corresponding actual test values. These can be found in Figures 6.9 through 6.14. On these ?45?-line? error plots, the center dark line represents the line of equality (zero error) between the estimated strength and the actual measured strength at each batch temperature. The other two sets of dotted lines represent ?10% and ?20% error. Points above the center line are overestimated strengths (unconservative), and points below the center line are underestimated strengths (conservative). The EoE values are evaluated by how far the points lie from the dark center line. The acceptable error tolerance for the purposes of this chapter will be ?10%, 124 which is a value that is commonly used in the concrete industry for the evaluation of cylinder strengths. The average absolute error of estimate was also computed for each batch of each mixture and for the total mixture. This value uses the same approach as Tank and Carino (1991) used to calculate the average absolute error. However, the average absolute error of estimate takes the average of the EoE values, neglecting signs. For clarification the equation is presented as follows: Average Absolute Error of Estimate (Abs. EoE) n EEo EoEAbs i n i ? =. Equation 6.3 where, Abs. EoE = Average absoute error of estimate (%), EoE = Error of estimate for ith test strength (%) n = Number of EoE values (6 for each batch; 18 for total abs. EoE). 125 Table 6.5: Error using NSM function based on ASTM C 1074 methods for the 20% F mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,510 1,540 30 2 1.0 2,660 2,360 -300 -11 2.0 3,380 3,420 40 1 7.0 5,120 5,350 230 4 14.1 5,930 6,160 230 4 Control 28.2 6,960 6,640 -320 -5 190 5 0.9 720 1,480 760 106 1.8 2,510 2,510 0 0 3.4 3,340 3,460 120 4 12.2 5,460 5,360 -100 -2 24.9 6,100 6,140 40 1 Cold 49.0 6,560 6,610 50 1 180 19 0.4 1,990 1,720 -270 -14 0.8 3,280 2,740 -540 -16 1.4 4,110 3,650 -460 -11 4.9 5,400 5,610 210 4 10.0 6,180 6,310 130 2 Hot 19.8 7,060 6,730 -330 -5 320 9 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.9: Estimated strength versus actual strength for 20% F mixture 126 Table 6.6: Error using NSM function based on ASTM C 1074 methods for the Type I - 0.41 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,030 1,060 30 3 1.0 2,700 2,670 -30 -1 2.1 4,060 3,910 -150 -4 7.1 4,900 5,300 400 8 14.0 5,690 5,690 0 0 Control 28.1 6,170 5,910 -260 -4 150 3 0.8 540 1,310 770 143 1.8 2,830 2,950 120 4 3.4 4,040 3,950 -90 -2 12.1 5,480 5,300 -180 -3 24.8 5,970 5,700 -270 -5 Cold 49.0 6,620 5,910 -710 -11 360 28 0.4 2,140 1,570 -570 -27 0.8 3,670 3,320 -350 -10 1.4 4,010 4,230 220 5 5.0 4,990 5,470 480 10 10.2 5,410 5,790 380 7 Hot 20.1 5,740 5,960 220 4 370 10 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.10: Estimated strength versus actual strength for the Type I - 0.41 mixture 127 Table 6.7: Error using NSM function based on ASTM C 1074 methods for the Type III - 0.37 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 4,280 4,410 130 3 1.0 6,530 6,290 -240 -4 2.1 7,970 7,850 -120 -2 7.1 9,110 9,730 620 7 14.3 10,100 10,270 170 2 Control 28.2 11,120 10,560 -560 -5 310 4 0.9 2,400 3,330 930 39 2.0 5,670 5,870 200 4 3.5 6,900 6,880 -20 0 12.6 9,060 9,130 70 1 25.1 10,270 9,890 -380 -4 Cold 49.4 10,520 10,340 -180 -2 300 8 0.3 2,860 3,650 790 28 0.8 6,400 7,200 800 13 1.4 6,960 8,390 1,430 21 5.0 8,030 10,040 2,010 25 10.0 8,610 10,430 1,820 21 Hot 19.8 9,030 10,590 1,560 17 1,400 21 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.11: Estimated strength versus actual strength for the Type III - 0.37 mixture 128 Table 6.8: Error using AM function based on ASTM C 1074 methods for the 20% F mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,510 1,540 30 2 1.0 2,660 2,360 -300 -11 2.0 3,380 3,420 40 1 7.0 5,120 5,360 240 5 14.1 5,930 6,160 230 4 Control 28.2 6,960 6,640 -320 -5 190 5 0.9 720 1,560 840 117 1.8 2,510 2,630 120 5 3.4 3,340 3,670 330 10 12.2 5,460 5,580 120 2 24.9 6,100 6,290 190 3 Cold 49.0 6,560 6,700 140 2 290 23 0.4 1,990 1,840 -150 -8 0.8 3,280 2,920 -360 -11 1.4 4,110 3,820 -290 -7 4.9 5,400 5,700 300 6 10.0 6,180 6,360 180 3 Hot 19.8 7,060 6,750 -310 -4 270 6 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.12: Estimated strength versus actual strength for the 20% F mixture 129 Table 6.9: Error using AM function based on ASTM C 1074 methods for the Type I - 0.41 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,030 1,060 30 3 1.0 2,700 2,670 -30 -1 2.1 4,060 3910 -150 -4 7.1 4,900 5,300 400 8 14.0 5,690 5,690 0 0 Control 28.1 6,170 5,910 -260 -4 150 3 0.8 540 1360 820 152 1.8 2,830 3,030 200 7 3.4 4,040 4,070 30 1 12.1 5,480 5,390 -90 -2 24.8 5,970 5,750 -220 -4 Cold 49.0 6,620 5,940 -680 -10 340 29 0.4 2,140 1,950 -190 -9 0.8 3,670 3,660 -10 0 1.4 4,010 4,460 450 11 5.0 4,990 5,550 560 11 10.2 5,410 5,840 430 8 Hot 20.1 5,740 5,980 240 4 310 7 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.13: Estimated strength versus actual strength for the Type I - 0.41 mixture 130 Table 6.10: Error using AM function based on ASTM C 1074 methods for the Type III - 0.37 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 4,280 4,400 120 3 1.0 6,530 6,310 -220 -3 2.1 7,970 7,830 -140 -2 7.1 9,110 9720 610 7 14.3 10,100 10,270 170 2 Control 28.2 11,120 10,570 -550 -5 300 4 0.9 2400 3,590 1,190 50 2.0 5,670 6,120 450 8 3.5 6,900 7,420 520 8 12.6 9,060 9,600 540 6 25.1 10,270 10,200 -70 -1 Cold 49.4 10,520 10,520 0 0 460 12 0.3 2,860 4,670 1,810 63 0.8 6,400 7,960 1,560 24 1.4 6,960 8,920 1,960 28 5.0 8,030 10,230 2,200 27 10.0 8,610 10,540 1,930 22 Hot 19.8 9,030 10,670 1,640 18 1,850 31 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.14: Estimated strength versus actual strength for the Type III - 0.37 mixture 131 For ease of comparison, the total absolute average errors and total average absolute error of estimates for each mixture and both maturity functions are reported in Table 6.11. The total average absolute errors are computed from Equation 6.1, using all strength calculations for a single mixture. The total average absolute error of estimate was computed from Equation 6.3. First, the accuracy of the Nurse-Saul maturity function using the ASTM C 1074 method for obtaining a mixture-specific datum temperature will be discussed for the three selected mixtures. Next, the accuracy of the Arrhenius maturity function using the ASTM C 1074 method for obtaining a mixture-specific activation energy will be discussed. Finally, the two maturity functions will be compared based on their total error values for all mixtures considered ,and the overall evaluation of accuracy of the ASTM C 1074 mixture-specific temperature sensitivity method will be discussed. 6.1.3.1 Accuracy of the Nurse-Saul Maturity Function 1. 20% F Mixture The accuracy of the strength estimations for the 20% F mixture using the NSM function and the mixture-specific datum temperature obtained from the ASTM C 1074 method is presented in Table 6.5 and Figure 6.9. From the table, it can be seen that the highest average absolute error came from estimating the hot batch?s strengths, with an AAE of 320 psi. Most of this error was attributed to underestimating the early-age strengths, which is conservative; estimated values were around 15% inaccurate at 0.4 and 0.8 days. The cold batch?s strengths were all estimated well, except for the earliest strength estimation. In fact, if the first point is dropped out of the AAE calculation, the 132 AAE goes from 180 psi to 60 psi. Also, because this mixture has little to no late-age strength loss or gain due to the crossover effect, all of the strengths from 7 days to 28 days equivalent age were estimated within ?5% of the actual values for all three batches. Also, the average absolute errors of estimates (abs. EoEs) for the hot and control batches were less than or equal to 9%. If the first strength estimation for the cold batch was neglected, its abs. EoE would be reduced from 19% to 2%. Table 6.11: Comparison of maturity functions for all mixtures, ASTM C1074 method Nurse-Saul Maturity (NSM) Arrhenius Maturity (AM) Mixture ID Total AAE (psi) Total Abs. EoE (%) Total AAE (psi) Total Abs. EoE (%) Function that Produced the Lowest Total Error Type I - 0.41 290 14 270 13 AM Type I - 0.44 380 16 290 14 AM Type I - 0.48 580 17 530 18 NSM 20% F 230 11 250 11 NSM 30% F 290 12 280 12 AM 20% C 250 6 220 7 NSM 30% C 280 11 270 11 AM 30% Slag 680 15 580 17 NSM 50% Slag 900 28 750 30 NSM Type III - 0.37 670 11 870 15 NSM Type III - 0.44 450 12 420 11 AM 70/20/10 - 0.37 700 13 870 19 NSM 70/20/10 - 0.44 730 23 660 26 NSM 133 2. Type I - 0.41 Mixture The accuracy of the Type I - 0.41 mixture using the NSM function and the mixture-specific datum temperature obtained from the ASTM C 1074 method can be seen in Table 6.6 and Figure 6.10. The AAE for both the hot and cold batches was around 360 psi. The first strength estimation for both the hot and cold batches was more than 20% inaccurate, where the cold batch?s strength was overestimated and the hot batch?s strength was underestimated. After the first test strength, all the estimated strengths for the hot and cold batches had 11% or less error. For the 7-day through 28-day equivalent age tests, the crossover effect was present and the hot strengths were overestimated and the cold strengths were underestimated. The average absolute error of estimate (Abs. EoE) for the hot batch was 10%. Again, neglecting the first strength estimation for the cold batch significantly reduces the abs. EoE, from 28% to 5%. 3. Type III - 0.37 Mixture The accuracy of the Type III - 0.37 mixture using the NSM function and the mixture-specific datum temperature obtained from the ASTM C 1074 method is shown in Table 6.7 and Figure 6.11. From this figure and table, it may be seen again that the first estimated strength for both the cold and hot batches was overestimated by more than 20%. Beyond that point, the cold batch?s strengths were estimated within ?4% of the actual strength values. Also, neglecting the first strength estimation, the remaining cold strength estimations had an AAE of 170 psi, a 2% abs. EoE, down from 300 psi for the full cold batch. All six of the hot batch?s strength estimations were overestimated by more than 10%. This shows how late-age strength losses due to high curing temperatures can adversely affect the accuracy of strength estimations using the maturity method. 134 Recall that in Chapter 5, Section 5.2.1, it was reported that this mixture had a 17% crossover factor for the hot batch versus the control batch. 6.1.3.2 Accuracy of the Arrhenius Maturity Function The accuracy of the Arrhenius maturity function using the mixture-specific activation energies, found from the ASTM C 1074 method, can be seen in Tables 6.8 through 6.10 and Figures 6.12 through 6.14. 1. 20% F Mixture The accuracy of the 20% F mixture using the AM function and the mixture- specific activation energy obtained from the ASTM C 1074 method can be seen in Table 6.8 and Figure 6.12. All of the strength estimations for the hot batch produced ?11% or less error. The first test strength for the cold batch was greatly overestimated, as was the case for the 20% F mixture when evaluated with the NSM function that was discussed earlier. However, the remaining strength estimations for the cold batch were all estimated within the ?10% error tolerance. 2. Type I - 0.41 Mixture The accuracy of the Type I - 0.41 mixture using the AM function and the mixture- specific activation energy obtained from the ASTM C 1074 method is shown in Table 6.9 and Figure 6.13. The strengths for the hot batch were all estimated within ?11% error. Again, the first strength estimation for the cold batch was greatly overestimated; however, the remaining strength estimations produced an abs. EoE of 5%. 135 3. Type III - 0.37 Mixture The accuracy of the Type III - 0.37 mixture using the AM function and the mixture-specific activation energy obtained from the ASTM C 1074 method can be found in Table 6.10 and Figure 6.14. Again, the first strength value for the cold batch was overestimated by more than 20%, while the remaining values for the cold batch were estimated within the ?10% error tolerance. The strength values for the hot batch were all overestimated by more than 10% and most by more than 20%. The average absolute error of estimate for the hot batch was 23%, which is caused by the presence of the crossover effect as discussed in Chapter 5, Section 5.2.1. Comparison of Maturity Functions for All Mixtures The total absolute average errors and total absolute average errors of estimate for all mixtures can be seen in Table 6.11. It is important to consider both the average error in strength units as well as the corresponding percentage error, because of the different strength-levels of the mixtures evaluated. Using the NSM function and the ASTM C 1074 method for obtaining mixture- specific datum temperatures, only 1 out of 13 mixtures produced total absolute average errors of estimate (total abs. EoE) of 10% or less. However, 9 out of 13 of the mixtures produced total abs. EoEs of 15% or less. As was observed from the three example mixtures discussed above, if the first strength estimation for the cold batch is neglected the total abs. EoEs would decrease significantly. Using the AM function and the ASTM method for obtaining mixture-specific activation energies, only 1 mixture produced a total abs. EoE of 10% or less. However, 8 out of 13 mixtures produced total abs. EoE of 136 15% or less and the same large early-age error trends were noticed from the cold batches. Overall, the Nurse-Saul maturity function produced the lower total abs. EoEs for 8 out 13 of the mixtures. However, it should be noted that for many of the mixtures (10 out of 13) the two functions were only 2% different in total abs. EoEs. The mixtures producing the worst total abs. EoEs, those producing more than 15% total abs. EoE, were the Type I - 0.48, 50% Slag, and 70/20/10 - 0.44 mixtures. These were all mixtures that had high (more than 20%) crossover factors for either their respective hot or cold batches. From Appendix D and the figures and tables presented above, it can be seen that the other mixtures that had high crossover factors for either the hot or cold batches, such as the 30% Slag mixture, both Type III mixtures, and the 70/20/10 - 0.37 mixture, also had high average absolute errors of estimate for their corresponding batches. For these mixtures, the crossover factors given in Table 5.3 are very similar to the late-age strength errors of estimate using the maturity method. For example, the crossover factor for the 70/20/10 - 0.44 hot batch was -31%; in other words, the hot batch?s late-age strengths were 31% less than the control batch?s late-age strengths. The last two strength estimations for the 70/20/10 - 0.44 hot batch using the NSM function were more than 40% overestimated. For the Type III - 0.37 mixture, the last three strength estimations were on average 21% overestimated, and the crossover factor for this batch was -17%. The Type III - 0.44 mixture had crossover factors of - 10% and 18% for the hot and cold batches respectively. The last three strength estimations for each of these batches were on average 10% overestimated and 16% underestimated for the hot and cold batches, respectively. Based on these late-age strength estimation errors, it may be hypothesized that the late-age strength values should 137 not be used in determining the accuracy of the maturity method to estimate concrete strength. As explained in Section 2.3.2, others (Kjellsen and Detwiler 1993; Jonasson 1985) also believe the maturity method should not be used to estimate late-age strengths. Kjellsen and Detwiler concluded that the maturity method is only valid up to 40% of the laboratory-cured 28-day strength, while Jonasson suggested 50% was the cutoff. Based on these facts, it is recommended that strength values should not be estimated after crossover has occurred. From the values presented in Table 5.3, the average time until crossover for the hot batches occurred at 10 days (excluding batches where crossover occurred after 28 days). The equivalent age for a 10-day hot batch strength, using the age conversion factor given in Section 4.3, is 14 days. Thus the 14 day and 28 day equivalent age test strengths will not be used to evaluate the accuracy of the maturity method in the following sections. It is also hypothesized that the rates of initial strength gain for the hyperbolic strength-age function (k-values), found in Chapter 5, were also affected by the long-term strength losses due to high curing temperatures. Therefore the least-squares regression analysis was repeated for the strength-age data only up to 7 days of equivalent age. If the k-values change, this would affect the mixture-specific temperature sensitivity values found using the ASTM method and, hence, the strength estimations. The next section describes evaluation of the accuracy of the maturity method using this ?Modified ASTM? method. 138 6.2 MODIFICATIONS TO THE ASTM C 1074 METHOD As was observed in the previous section, the crossover effect can cause severe errors when estimating late-age strengths using the maturity method. Therefore, it is proposed that the maturity method be used only to estimate strengths up to 7 days of equivalent age. Accordingly, the k-value method of determining the mixture-specific temperature sensitivity values should be based only on the strength-age data up to 7 days of equivalent age. By neglecting the later-age strengths, the influence of the crossover effect on early-age strength estimates should be reduced. The use of only early-age strength values should change the k-values, or the rates of initial reaction. Most notably, the k-value for hot batches should increase for mixtures with crossover, because the plateau of strength gain at the later-ages will not affect the determination of the k-value. An example of this rate increase and subsequent temperature sensitivity change is shown in Figure 6.15 for the Type I - 0.41 mixture. In the figure, the diamond data points represent the new k-values found by only considering strength data up to 7 days of equivalent age. The square data points represent the k-values found when all the strength data up to 28 days of equivalent age are considered. The figure shows that the cold k- value did not change significantly; however, the control and hot batch?s k-values increased greatly, bringing the mixture-specific datum temperature up to 27?F (-3?C). 139 y = 0.00179x - 0.04832 R2 = 0.99176 y = 0.0011x - 0.0246 R2 = 0.9991 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 20 30 40 50 60 70 80 90 100 110 Temperature (?F) Rate Constant (1/hr) te up to 7 d All Data Figure 6.15: Change in rate values and datum temperature for Type I - 0.41 mixture Throughout this section, this new method of determining mixture-specific temperature sensitivity values is referred to as the ?Modified ASTM? method, as calculation of the temperature sensitivity values remains the same as documented in ASTM C 1074; however, only data up to an equivalent age of 7 days are considered. Again, S-M relationships are shown for the three example mixtures. Next, temperature sensitivity values are given and compared to the ASTM method and previously published values. Strength estimation errors determined with both maturity functions using the Modified ASTM method will be reported and compared to those obtained from the ASTM C 1074 method. The results from the least-squares regression analysis of the strength-age data for all mixtures evaluated only up to 7 days of equivalent age are reported in Table 6.12. To = 27?F 140 Table 6.12: Regression values for strength-age curves up to 7 days of equivalent age Mixture ID Batch ID to (hr) Su (psi) k (hr-1) 3.6 5,270 0.126 9.7 5,370 0.073 Type I - 0.41 Hot Control Cold 16.1 6,070 0.032 1.2 5,090 0.054 7.4 4,850 0.038 Type I - 0.44 Hot Control Cold 16.6 5,350 0.021 2.7 4,670 0.050 6.3 4,970 0.027 Type I - 0.48 Hot Control Cold 17.5 5,810 0.015 3.5 6,040 0.074 2.8 6,110 0.030 20% F Hot Control Cold 12.5 6,530 0.017 4.3 5,610 0.063 4.4 5,440 0.028 30% F Hot Control Cold 13.7 6,270 0.014 4.9 5,790 0.090 7.1 6,370 0.033 20% C Hot Control Cold 12.8 7,190 0.011 7.0 6,880 0.051 9.7 6,520 0.025 30% C Hot Control Cold 21.4 5,880 0.014 4.5 5,900 0.077 2.7 7,110 0.025 30% Slag Hot Control Cold 15.4 6,200 0.017 2.8 8,480 0.021 5.3 8,420 0.013 50% Slag Hot Control Cold 1.8 10,740 0.003 4.4 8,300 0.207 8.3 9,510 0.132 Type III - 0.37 Hot Control Cold 13.9 9,820 0.038 0.0 5,380 0.089 7.6 5,500 0.066 Type III - 0.44 Hot Control Cold 16.6 6,660 0.020 0.0 9,190 0.035 5.2 9,130 0.031 70/20/10 - 0.37 Hot Control Cold 12.3 8,420 0.018 0.0 6,030 0.025 4.8 6,180 0.018 70/20/10 - 0.44 Hot Control Cold 19.1 5,720 0.014 141 6.2.1 Nurse-Saul Maturity Function The newly found k-values given in Table 6.12 were used to find the appropriate datum temperatures for each mixture. The new datum temperatures are shown in Table 6.13, along with the values found previously in Section 6.1.1. There are various differences in the datum temperatures found from the Modified ASTM method and ASTM C 1074 method. Table 6.13: Datum temperatures obtained from different methods Modified ASTM ASTM C 1074 Mixture ID Datum Temperature To (?F (?C)) Datum Temperature To (?F (?C)) Type I - 0.41 27 (-3) 23 (-5) Type I - 0.44 17 (-9) 35 (2) Type I - 0.48 28 (-2) 39 (4) 20% F 36 (2) 23 (-5) 30% F 39 (4) 19 (-7) 20% C 46 (8)* 37 (3) 30% C 35 (2) 42 (5)* 30% Slag 40 (5) 46 (8)* 50% Slag 25 (-4) 47 (9)* Type III - 0.37 33 (1) 38 (3) Type III - 0.44 29 (-2) 37 (3) 70/20/10 - 0.37 -15 (-26) 39 (4) 70/20/10 - 0.44 -17 (-27) 43 (6)* * Value higher than some temperatures reached during cold temperature history The Type I - 0.41 mixture?s datum temperature increased from 23?F (-5?C) to 27?F (-3?C), as was shown in Figure 6.15. The straight Type I and Type III cement mixtures? datum temperatures have a much smaller spread, from 27?F to 33?F, excluding 142 the Type I - 0.44 mixture. Both Class F fly ash mixtures increased in datum temperature, while both GGBF Slag mixtures decreased. The biggest difference is noticed for the ternary blend mixtures. These mixtures contain silica fume, which has a very high rate of reaction when used in concrete (Holland 2005). Recall that Saul (1951) defined the datum temperature as the temperature at which a concrete ceases to gain strength, after it has enough time to set. With that in mind, these low datum temperatures may indicate that once the silica fume begins to react, it is very insensitive to the curing temperature. Referring back to Table 6.2, the datum temperature values found from the Modified ASTM method are much different than those Carino and Tank (1992) found, but this is to be expected as only early-age data was considered and the rate constants changed in this study. Next, the strength-maturity (S-M) relationships are shown for the Nurse-Saul maturity function with the mixture-specific datum temperatures found from the Modified ASTM method. These plots along with the measured data can be seen in Figures 6.16, 6.17, and 6.18. These figures may be compared to Figures 6.2, 6.3, and 6.4; however, note that the horizontal scales have been changed to only include data up to 7 days of equivalent age. The remaining S-M relationships for the Nurse-Saul Function, using the Modified ASTM method, may be found in Appendix E. The accuracy of strength estimations using this method are discussed at the end of this section. 143 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.16: Strength-maturity plot, Modified ASTM method for 20% F mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.17: Strength-maturity plot, Modified ASTM method for Type I - 0.41 To = 36?F To = 27?F 144 0 2,000 4,000 6,000 8,000 10,000 12,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Realtionship 73 ?F Cold Hot Figure 6.18: Strength-maturity plot, Modified ASTM method for Type III - 0.37 mixture 6.2.2 Arrhenius Maturity Function Next, the Modified ASTM method of determining mixture-specific temperature sensitivity values was applied to the Arrhenius maturity function. New activation energies, based on the k-values given in Table 6.12, were found and are reported in Table 6.14. The trends for the new activation energies are similar to those found earlier with the Modified ASTM datum temperatures. The Type - 0.41 activation energy increased, and the Type I - 0.44 and Type I - 0.48 activation energies decreased. The Type I straight cement mixtures are closer together, around 30,000 J/mol. However, the straight Type III mixtures? activation energies remained high, averaging approximately 40,000 J/mol. Both Class F fly ash mixtures increased activation energies, just as the Modified ASTM datum temperatures increased for these mixtures. Also, both ternary mixtures had an To = 33?F 145 extreme reduction in activation energy, from about 45,000 J/mol to 15,000 J/mol, which suggests that the ternary mixtures have very low temperature sensitivity, as was discussed with the datum temperatures previously. Table 6.14: Activation energies based obtained from different methods Modified ASTM ASTM C1074 Mixture ID Activation Energy, E (J/mol) Activation Energy, E (J/mol) Type I - 0.41 33,100 28,600 Type I - 0.44 25,000 34,800 Type I - 0.48 30,800 42,300 20% F 37,000 25,800 30% F 40,200 23,200 20% C 54,100 37,800 30% C 35,300 45,100 30% Slag 39,000 55,700 50% Slag 28,600 61,500 Type III - 0.37 42,000 41,400 Type III - 0.44 38,100 42,900 70/20/10 - 0.37 16,700 42,600 70/20/10 - 0.44 14,800 47,300 The S-M relationships for the three example mixtures, using the Arrhenius maturity function and mixture-specific activation energies found using the Modified ASTM method, are shown in Figures 6.19, 6.20, and 6.21. These plots may be compared to Figures 6.6, 6.7, and 6.8, respectively; however, the horizontal scales were changed to only include data up to 7 days of equivalent age. The remaining S-M relationships for the Arrhenius Maturity Function, using the Modified ASTM method, may be found in 146 Appendix F. The accuracy of strength estimations using the NSM and AM functions using the Modified ASTM method will be discussed in the next section. 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 50 100 150 200 250 300 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.19: Strength-maturity plot, Modified ASTM method for 20% F mixture E = 37,000 J/mol 147 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 300 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.20: Strength-maturity plot, Modified ASTM method for Type I - 0.41 mixture 0 2,000 4,000 6,000 8,000 10,000 12,000 0 50 100 150 200 250 300 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure 6.21: Strength-maturity plot, Modified ASTM method for Type III - 0.37 mixture E = 33,100 J/mol E = 42,000 J/mol 148 6.2.3 Accuracy of Maturity Functions Based on the Modified ASTM Method The errors of the three example mixtures using the Modified ASTM method are shown in Tables 6.15 through 6.17 using the NSM function and Tables 6.18 to 6.20 using the AM function. The remaining mixtures? error tables can be found in Appendix G. In these tables, the errors from the ASTM C 1074 method are shown for ease of comparison. The AAEs for the ASTM C 1074 method were recalculated to include strengths only up to 7 days of equivalent age. The 45?-line error plot for each mixture using the ASTM C 1074 and Modified ASTM methods can be seen on each page following the corresponding error tables. 149 Table 6.15: Error using NSM function based on Modified ASTM method for 20% F mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,510 1,540 30 2 1,570 60 4 1.0 2,660 2,360 -300 -11 2,500 -160 -6 2.0 3,380 3,420 40 1 3,510 130 4 Control 7.0 5,120 5,350 230 4 150 5 5,090 -30 -1 100 4 0.9 720 1,480 760 106 1,260 540 75 1.8 2,510 2,510 0 0 2,290 -220 -9 3.4 3,340 3,460 120 4 3,100 -240 -7 Cold 12.2 5,460 5,360 -100 -2 250 28 4,700 -760 -14 440 26 0.4 1,990 1,720 -270 -14 1,960 -30 -2 0.8 3,280 2,740 -540 -16 3,090 -190 -6 1.4 4,110 3,650 -460 -11 3,930 -180 -4 Hot 4.9 5,400 5,610 210 4 370 11 5,440 40 1 110 3 150 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.22: Error plot for the 20% F mixture, Modified ASTM method, NSM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.23: Error plot for the 20% F mixture, ASTM C 1074 method, NSM 151 Table 6.16: Error using NSM function based on Modified ASTM method for Type I - 0.41 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,030 1,060 30 3 1,000 -30 -3 1.0 2,700 2,670 -30 -1 2,810 110 4 2.1 4,060 3,910 -150 -4 3,920 -140 -3 Control 7.1 4,900 5,300 400 8 150 4 4,960 60 1 85 3 0.8 540 1,310 770 143 1,210 670 124 1.8 2,830 2,950 120 4 2,980 150 5 3.4 4,040 3,950 -90 -2 3,860 -180 -4 Cold 12.1 5,480 5,300 -180 -3 290 38 4,910 -570 -10 393 36 0.4 2,140 1,570 -570 -27 1,740 -400 -19 0.8 3,670 3,320 -350 -10 3,490 -180 -5 1.4 4,010 4,230 220 5 4,220 210 5 Hot 5.0 4,990 5,470 480 10 410 13 5,090 100 2 223 8 152 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.24: Error plot for the Type I - 0.41 mixture, Modified ASTM method, NSM 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.25: Error plot for the Type I - 0.41 mixture, ASTM C 1074 method, NSM 153 Table 6.17: Error using NSM function based on Modified ASTM method for Type III - 0. 37 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 4,280 4,410 130 3 4,270 -10 0 1.0 6,530 6,290 -240 -4 6,560 30 0 2.1 7,970 7,850 -120 -2 7,930 -40 -1 Control 7.1 9,110 9,730 620 7 280 4 9,120 10 0 20 0 0.9 2,400 3,330 930 39 3,100 700 29 2.0 5,670 5,870 200 4 6,500 830 15 3.5 6,900 6,880 -20 0 7,500 600 9 Cold 12.6 9,060 9,130 70 1 310 11 8,950 -110 -1 560 13 0.3 2,860 3,650 790 28 2,720 -140 -5 0.8 6,400 7,200 800 13 7,300 900 14 1.4 6,960 8,390 1,430 21 8,230 1,270 18 Hot 5.0 8,030 10,040 2,010 25 1260 21 9,260 1,230 15 890 13 154 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.26: Error plot for the Type III - 0.37 mixture, Modified ASTM method, NSM 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.27: Error plot for the Type III - 0.37 mixture, ASTM C 1074 method, NSM 155 Table 6.18: Error using AM function based on Modified ASTM method for 20% F mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,510 1,540 30 2 1,570 60 4 1.0 2,660 2,360 -300 -11 2,500 -160 -6 2.0 3,380 3,420 40 1 3,510 130 4 Control 7.0 5,120 5,360 240 5 150 5 5,090 -30 -1 95 4 0.9 720 1,560 840 117 1,390 670 93 1.8 2,510 2,630 120 5 2,490 -20 -1 3.4 3,340 3,670 330 10 3,450 110 3 Cold 12.2 5,460 5,580 120 2 350 33 5,100 -360 -7 290 26 0.4 1,990 1,840 -150 -8 2,320 330 17 0.8 3,280 2,920 -360 -11 3,510 230 7 1.4 4,110 3,820 -290 -7 4,270 160 4 Hot 4.9 5,400 5,700 300 6 280 8 5,580 180 3 225 8 156 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.28: Error plot for the 20% F mixture, Modified ASTM method, AM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.29: Error plot for the 20% F mixture, ASTM C 1074 method, AM 157 Table 6.19: Error using AM function based on Modified ASTM method for Type I - 0.41 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,030 1,060 30 3 1,000 -30 -3 1.0 2,700 2,670 -30 -1 2,820 120 4 2.1 4,060 3,910 -150 -4 3,910 -150 -4 Control 7.1 4,900 5,300 400 8 150 4 4,960 60 1 90 3 0.8 540 1,360 820 152 1,260 720 133 1.8 2,830 3,030 200 7 3,040 210 7 3.4 4,040 4,070 30 1 3,970 -70 -2 Cold 12.1 5,480 5,390 -90 -2 290 40 4,990 -490 -9 370 38 0.4 2,140 1,950 -190 -9 2,320 180 8 0.8 3,670 3,660 -10 0 3,880 210 6 1.4 4,010 4,460 450 11 4,460 450 11 Hot 5.0 4,990 5,550 560 11 300 8 5,170 180 4 260 7 158 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.30: Error plot for the Type I - 0.41 mixture, Modified ASTM method, AM 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.31: Error plot for the Type I - 0.41 mixture, ASTM C 1074 method, AM 159 Table 6.20: Error using AM function based on Modified ASTM method for Type III - 0.37 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 4,280 4,400 120 3 4,270 -10 0 1.0 6,530 6,310 -220 -3 6,580 50 1 2.1 7,970 7,830 -140 -2 7,900 -70 -1 Control 7.1 9,110 9,720 610 7 270 4 9,130 20 0 40 1 0.9 2,400 3,590 1,190 50 2,870 470 20 2.0 5,670 6,120 450 8 6,360 690 12 3.5 6,900 7,420 520 8 7,560 660 10 Cold 12.6 9,060 9,600 540 6 680 18 9,050 -10 0 460 10 0.3 2,860 4,670 1,810 63 4,720 1,860 65 0.8 6,400 7,960 1,560 24 8,020 1,620 25 1.4 6,960 8,920 1,960 28 8,670 1,710 25 Hot 5.0 8,030 10,230 2,200 27 1880 36 9,410 1,380 17 1640 33 160 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.32: Error plot for the Type III - 0.37 mixture, Modified ASTM method, AM 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.33: Error plot for the Type III - 0.37 mixture, ASTM C 1074 method, AM 161 Accuracy of the Nurse-Saul Maturity Function 1. 20% F Mixture From Table 6.15, it can be seen that the Modified ASTM method did help reduce the hot batch?s error from an abs. EoE of 11% to 3%. The first test strength for the cold batch was still highly overestimated and the Modified ASTM method made the other three strength estimates for the cold batch worse than the normal ASTM C 1074 method. 2. Type I - 0.41 Mixture The error for the Type I - 0.41 mixture using the NSM function and the Modified ASTM method can be seen in Table 6.16. Again, the Modified ASTM method reduced the average percent error for the hot batch, down from 13% using the ASTM C 1074 method to 8% using the Modified ASTM method. The strength estimation for the 0.4 day strength was still overestimated by 19%, but the remaining strengths were estimated within ?5% of the actual test values. The Modified method decreased the average percent error for the cold batch from 38% to 36%, but increased the error for the 12-day strength estimation from -3% to -10%. 3. Type III - 0.37 Miixture The strength estimation errors using the NSM function and the Modified ASTM method can be seen in Table 6.17. As before, the Modified ASTM method reduced the average percent error of estimate, from 21% to 13%, for the hot batch. However, it slightly increased the error of estimate from 11% to 13% for the cold batch. The Modified method lowered all of the strength estimations for the hot batch to 18% error or less. 162 Accuracy of the Arrhenius Maturity Function 1. 20% F Mixture The strength estimation errors for the 20% F mixture using the Arrhenius maturity function and the mixture-specific activation energy found from the Modified ASTM method can be seen in Table 6.18. This method reduced the percent error of estimate for all three batches. Aside from high errors for the first strength estimation for the hot and cold batches, the remaining estimations were all within ?7% of the actual test value. 2. Type I - 0.41 Mixture The strength estimation errors using the AM function with the mixture-specific activation energy found from the Modified ASTM method are shown in Table 6.19. There was a slight decrease in abs. EoE for estimating the hot batch?s strengths. The Modified ASTM method slightly decreased the average percent error for the cold batch from 40% to 38%. Again, the earliest test age is the bulk of the error or estimations for the cold batch. 3. Type III - 0.37 Mixture The strength estimation errors for the Type III - 0.37 mixture using the AM function and the mixture-specific activation energy found from the Modified ASTM method are shown in Table 6.20. The Modified method reduced the average percent errors for all three batches. The earliest strength estimation for the cold batch was about 40% more accurate using the Modified ASTM method versus the ASTM C 1074 method; however, both methods provide reasonable results for the cold curing condition thereafter. The average percent error for the hot batch was reduced slightly from 36% to 163 33% by using the Modified ASTM method, but still none of the estimations were within ?10% of the actual values. Comparison of Functions for All Mixtures Table 6.21 shows the total average absolute percent errors and average absolute errors for all mixtures using the NSM and AM functions with the mixture-specific temperature sensitivity values found using the Modified ASTM method. From the table it can be seen that the Nurse-Saul maturity function produced the lowest percent error of estimations for 7 out of the 13 mixtures considered. However, 7 out the 13 mixtures have only 2% or less difference in total error between the two functions. When using the NSM function, the Modified ASTM method produced less total error than the ASTM C 1074 method for all mixtures except the 20% C, Type III - 0.44, and ternary-blend mixtures. When using the AM function, the Modified ASTM method produced less error than the ASTM C 1074 method for all mixtures except the 20% C and ternary-blend mixtures. Because of the increased accuracy of the estimated strengths that resulted from using the maturity method only up to 7 days equivalent age, the remaining analysis reported in this chapter only considered the accuracy of the maturity method based on data up to 7 days equivalent age. 164 Table 6.21: Comparison of maturity methods for all mixtures, Modified ASTM method NSM AM Error lower than ASTM C 1074? Mixture ID Total AAE (psi) Total abs. EoE (%) Total AAE (psi) Total abs. EoE (%) Function that Produced Lowest Error NSM AM Type I - 0.41 230 16 240 16 NSM Type I - 0.44 230 19 170 17 AM Type I - 0.48 210 14 200 13 AM 20% F 220 11 200 12 NSM 30% F 250 10 210 12 NSM 20% C 640 24 450 16 AM 30% C 150 7 210 13 NSM 30% Slag 250 10 390 18 NSM 50% Slag 160 14 250 17 NSM Type III - 0.37 490 9 710 15 NSM Type III - 0.44 310 15 275 11 AM 70/20/10 - 0.37 540 23 530 22 AM 70/20/10 - 0.44 320 35 310 34 AM = NO, = YES 6.3 SIMPLIFIED MATURITY APPROACH After some methods of finding mixture-specific temperature sensitivity values were evaluated, researchers investigated the difference in using constant, commonly used values for all mixtures, relative to using mixture-specific values for each mixture. This would obviously be a more practical approach to using the maturity method for actual construction projects. As mentioned previously, the remainder of this chapter reports only evaluations using data up to 7 days of equivalent age. 165 To choose the constant temperature sensitivity values, an age conversion factor versus temperature plot was constructed, as explained in Section 2.4.1. The procedure to find age conversion factors for a particular mixture was also explained in Section 2.4.1. In summary, to get the age conversion factor (ACF) for each mixture, the rate constant (k-value) at each temperature is divided by the rate constant at the reference temperature, taken as 73?F (23?C) in this study. This can be performed easily with the linear trendline that was used earlier in order to find the best-fit datum temperatures or activation energies for each mixture. An illustration of this process for the Type I - 0.41 mixture is provided in Figure 6.34. The value of the rate constant at the reference temperature is shown to be 0.083 hr -1. The ACFs for the hot, control, and cold batches were found by dividing their respective rate constants by 0.083 hr -1. As was explained in Section 2.4.1, to find the rate constant at the reference temperature, Carino (1991) compared the linear trendline used to find the mixture- specific datum temperatures to the trendline used to find the mixture-specific activation energies. The line that fit the data best was the line he used to determine the k-value at the reference temperature. Carino used the R2 values of the two trendlines to determine which equation to use. This procedure was also used in this study. The R2 values for the trendlines of all mixtures are reported in Table 6.22. Also the corresponding rate constant at the reference temperature, obtained from the equation with the highest R2, for each mixture is given. The age conversion factors were computed for all mixtures, and the ACF versus temperature plot for all mixtures is shown in Figure 6.35. The ACFs for the NSM function, defined by Equation 2.7, using To = 14?F (-10?C) and 32?F (0?C) are the two 166 straight lines shown on the plot. The ACFs for the AM function, also defined in Equation 2.7, using E = 40,000 J/mol and 25,000 J/mol are the two non-linear relationships shown. This type of plot is the same as those shown in Figures 2.16, 2.19, and 2.20. y = 0.00179x - 0.04832 R2 = 0.99176 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 30 40 50 60 70 80 90 100 110 Temperature (?F) Rate Constant (1/hr) te up to 7 d Figure 6.34: Obtaining age conversion factors for Type I - 0.41 mixture When examining an age conversion factor versus temperature plot, the further an age conversion factor, for a certain mixture, deviates from an age conversion factor of 1.0, the greater the temperature sensitivity requirements are for that mixture. Therefore, mixtures with behavior similar to the ternary-blends, with age conversion factors of approximately 1.25 and 0.6, for the hot and cold batches, respectively, would have low temperature sensitivity requirements. The maturity function that best models a concrete?s temperature sensitivity behavior at different temperatures, theoretically, will be the function that produces the least amount of strength estimation errors. From Figure 6.35 Tref = 73?F kRef. = 0.083 ACFcontrol = 0.073/0.083 ACFcontrol = 0.88 ACFcold = 0.032/0.083 ACFcold = 0.39 ACFhot = 0.126/0.083 ACFhot = 1.53 167 several other temperature sensitivity trends may be seen. Several of the mixtures, such as the Type III mixtures and the Type I - 0.41 mixture, require a high temperature sensitivity at low curing temperatures and a lower temperature sensitivity at high curing temperatures. The behavior provided by a linear ACF function, like the Nurse-Saul maturity function, would be ideal for these mixtures. Other mixtures, like the 30% Slag, 20% F, 20% C, and 30% C, require high temperature sensitivity values at high and low curing temperatures. The behavior provided by a non-linear ACF function, like the Arrhenius maturity function, would be better suited for these mixtures. 168 Table 6.22: R2 values and rate constants at the reference temperature for all mixtures R2 from trendline of: Mixture ID k versus T(?F) plot Ln(k) versus T(K) plot k at 73?F (1/hr) Type I - 0.41 0.992 0.936 0.083 Type I - 0.44 0.987 0.954 0.038 Type I - 0.48 0.988 0.998 0.028 20% F 0.964 0.998 0.036 30% F 0.998 0.980 0.035 20% C 0.991 0.977 0.043 30% C 0.993 0.996 0.027 30% Slag 0.861 0.925 0.032 50% Slag 0.995 0.960 0.014 Type III - 0.37 0.972 0.888 0.129 Type III - 0.44 0.937 0.868 0.059 70/20/10 - 0.37 0.829 0.806 0.029 70/20/10 - 0.44 0.997 0.987 0.019 169 0.0 0.5 1.0 1.5 2.0 2.5 34 44 54 64 74 84 94 104 Average Curing Temperature (?F) Age Conversion Factor To = 14?F To = 32?F E = 40,000 J/mol E = 25,000 J/mol Type I - 0.41 Type I - 0.44 Type I - 0.48 20% C 30% C 20% F 30% F 30% Slag 50% Slag Type III - 0.37 Type III - 0.44 70/20/10 - 0.37 70/20/10 - 0.44 Figure 6.35: Age conversion factors for all mixtures AM (E = 40,000 J/mol) AM (E = 25,000 J/mol) NSM (To = 14?F [-10?C]) NSM (To = 32?F [0?C]) 170 Constant temperature sensitivity values were chosen after visual inspection of the results in Figure 6.35. A datum temperature of 32?F (0?C) appears to accurately model the temperature sensitivity of mixtures with behavior like the Type I - 0.41 and Type III mixtures; however, 14?F (-10?C) was also analyzed as it is another commonly used value (as discussed in Section 2.2.4). An activation energy of 25,000 J/mol models the average of the hot batch?s data well, while an activation energy of 40,000 J/mol models specific mixtures like the 30% Slag better. Both maturity functions were analyzed using the appropriate two constant temperature sensitivity values mentioned above. Accuracy was compared to the Modified ASTM method to determine if there is any loss/gain in accuracy and, if so, to what degree. 6.3.1 Nurse-Saul Maturity Function First, the compressive strength data was analyzed using the Nurse-Saul maturity (NSM) function. The results for the three example mixtures are summarized in Tables 6.23 through 6.25. Figures 6.36 to 6.41 also show the 45?-line error plots associated with the NSM function using the constant datum temperatures. The remaining error tables and accompanying 45?-line error plots can be found in Appendix H. For the three example mixtures, the NSM function using 32?F (0?C) as the datum temperature produces less error than using 14?F (-10?C) as the datum temperature. The total absolute errors and percent errors for all thirteen mixtures can be found in Table 6.26. For all thirteen mixtures, the NSM function using 32?F (0?C) as the datum temperature produced lower total AAE than using 14?F (-10?C) as the datum temperature. For 11 out of the 13 mixtures the NSM function use of 32?F (0?C) as the datum temperature produced lower 171 average percent errors for both the hot and cold batches. The two mixtures for which this did not occur were the ternary-blend mixtures. Recall that the mixture-specific datum temperatures for the ternary-blend mixtures using the Modified ASTM method averaged -16?F (-26?C). For these mixtures using 14?F (-10?C) as the datum temperature produced less error for the hot batch?s strength estimations, while the 32?F (0?C) datum temperature produced lower errors for the cold batch?s strength estimations. This suggests that the ternary-blend mixtures have a very low temperature sensitivity for the hot batch and a higher temperature sensitivity for the cold batch. These results suggest that some mixtures may have different temperature sensitivity values for the hot- and cold-cured batches and that a constant mixture-specific temperature sensitivity value may not always be appropriate for all curing temperatures. Table 6.26 also shows that using a datum temperature of 32?F (0?C) produced less total average error than using mixture-specific datum temperatures found from the Modified ASTM method for 8 out of the 13 mixtures evaluated. For 6 out of the 13 mixtures, using 32?F (0?C) as the datum temperature produced total average absolute errors only 2% different than those produced using mixture-specific datum temperatures found from the Modified ASTM method. From these results it seems that the 32?F (0?C) datum temperature may not be the ideal value for all curing temperatures of all mixtures; however, the average error resulting from using this value as opposed to mixture-specific values is approximately the same if not less. . 172 Table 6.23: Error using NSM function based on simplified method for 20% F mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,510 1,660 150 10 1,580 70 5 1.0 2,660 2,460 -200 -8 2,490 -170 -6 2.0 3,380 3,460 80 2 3,510 130 4 Control 7.0 5,120 5,110 -10 0 110 5 5,090 -30 -1 100 4 0.9 720 1,680 960 133 1,350 630 88 1.8 2,510 2,730 220 9 2,410 -100 -4 3.4 3,340 3,650 310 9 3,270 -70 -2 Cold 12.2 5,460 5,240 -220 -4 430 39 4,870 -590 -11 350 26 0.4 1,990 1,750 -240 -12 1,900 -90 -5 0.8 3,280 2,720 -560 -17 3,010 -270 -8 1.4 4,110 3,560 -550 -13 3,860 -250 -6 Hot 4.9 5,400 5,240 -160 -3 380 11 5,380 -20 0 160 5 173 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.36: Error plot for the 20% F mixture, simplified method, To = 14?F 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.37: Error plot for the 20% F mixture, simplified method, To = 32?F 174 Table 6.24: Error using NSM function based on simplified method for Type I - 0.41 mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,030 1,000 -30 -3 1,000 -30 -3 1.0 2,700 2,800 100 4 2,820 120 4 2.1 4,060 3,930 -130 -3 3,910 -150 -4 Control 7.1 4,900 4,950 50 1 80 3 4,960 60 1 90 3 0.8 540 1,470 930 172 1,050 510 94 1.8 2,830 3,220 390 14 2,830 0 0 3.4 4,040 4,080 40 1 3,700 -340 -8 Cold 12.1 5,480 5,000 -480 -9 460 49 4,820 -660 -12 380 29 0.4 2,140 1,450 -690 -32 1,880 -260 -12 0.8 3,670 3,310 -360 -10 3,580 -90 -2 1.4 4,010 4,100 90 2 4,290 280 7 Hot 5.0 4,990 5,030 40 1 300 11 5,120 130 3 190 6 175 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.38: Error plot for the Type I - 0.41 mixture, simplified method, To = 14?F 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.39: Error plot for the Type I - 0.41 mixture, simplified method, To = 32?F 176 Table 6.25: Error using NSM function based on simplified method for Type III - 0.37 mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 4,280 4,280 0 0 4,270 -10 0 1.0 6,530 6,540 10 0 6,560 30 0 2.1 7,970 7,960 -10 0 7,930 -40 -1 Control 7.1 9,110 9,110 0 0 10 0 9,120 10 0 20 0 0.9 2,400 4,510 2,110 88 3,280 880 37 2.0 5,670 7,190 1,520 27 6,590 920 16 3.5 6,900 8,100 1,200 17 7,580 680 10 Cold 12.6 9,060 9,160 100 1 1230 33 8,980 -80 -1 640 16 0.3 2,860 1,710 -1,150 -40 2,630 -230 -8 0.8 6,400 6,960 560 9 7,260 860 13 1.4 6,960 8,020 1,060 15 8,210 1,250 18 Hot 5.0 8,030 9,160 1,130 14 980 20 9,250 1,220 15 890 14 177 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.40: Error plot for the Type III - 0.37 mixture, simplified method, To = 14?F 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.41: Error plot for the Type III - 0.37 mixture, simplified method, To = 32?F 178 Table 6.26: Comparison of NSM functions for all mixtures, simplified method To = 14?F (-10?C) To = 32?F (0?C) Modified ASTM Mixture ID Total AAE (psi) Total abs. EoE (%) Total AAE (psi) Total abs. EoE (%) Total AAE (psi) Total abs. EoE (%) Function that Produced Lowest Error Type I - 0.41 280 21 220 13 230 16 To = 32?F Type I - 0.44 240 20 200 12 230 19 To = 32?F Type I - 0.48 230 20 260 12 210 14 To = 32?F 20% F 310 18 200 12 220 11 Modified 30% F 290 16 220 10 250 10 Modified 20% C 330 16 210 8 640 24 To = 32?F 30% C 440 26 190 9 150 7 Modified 30% Slag 400 23 250 14 250 10 Modified 50% Slag 300 23 70 6 160 14 To = 32?F Type III - 0.37 740 18 520 10 490 9 Modified Type III - 0.44 400 23 300 14 310 15 To = 32?F 70/20/10 - 0.37 450 18 310 11 540 23 To = 32?F 70/20/10 - 0.44 230 26 200 17 320 35 To = 32?F 6.3.2 Arrhenius Maturity Function Next, the mixtures will be analyzed with activation energies of 25,000 J/mol and 40,000 J/mol using the Arrhenius maturity function. The errors using these activation energies are shown in Tables 6.27 through 6.29. The 45?-line error plots are also shown in Figures 6.42 to 6.47. The remaining error plots and accompanying 45?-line error plots can be found in Appendix I. For the 20% F mixture, the higher activation energy 179 produced lower errors for the cold batch and the lower activation energy produced lower errors for the hot batch. However, from Figure 6.35 one would think that the higher activation energy would produce the lower errors for the hot and cold batches. This suggests that the rate constants computed for the Modified ASTM method, for this mixture, does not accurately describe the actual temperature sensitivity requirements of this mixture. It may be seen in Table 6.12 that the k-value is not the only parameter affected by curing temperature. Both the Su and to values are affected by the curing temperature, yet the procedure to determine the mixture-specific temperature sensitivity values only accounts for changes in the rate constant, k. This fundamental flaw could be to blame for such obscure trends, like that seen from the 20% F mixture. For the Type I - 0.41 and Type III - 0.37 mixtures, the higher activation energy also produced lower errors for the cold batches, while the lower activation energy produced lower errors for the hot batches. This, however, does agree with trends observed from the ACF versus temperature plot. For the 10 remaining mixtures, eight had the same activation energy trend as the Type I - 0.41 and Type III - 0.37 mixtures. The higher activation energy produced lower errors for the lower curing temperature mixtures and the lower activation energy produced lower errors for the higher curing temperature mixtures. The total average errors and percent errors for all mixtures are reported in Table 6.30. Nine out of the 13 mixtures considered had lower total average errors using 40,000 J/mol as an activation energy. This concurs with the recommendation for activation energy outlined in ASTM C 1074. However, from the error tables and the conclusions stated above, it is evident that the best Arrhenius function would use a lower activation 180 energy value at high curing temperatures and a higher value at lower curing temperatures. This has been observed and applied in previous studies, as seen with the FHP model (see Section 2.4.2). This occurrence is the motivation for the last analysis presented in this chapter. 181 Table 6.27: Error using AM function based on simplified method for 20% F mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AA E (psi) Abs . EoE (%) 0.6 1,510 1,570 60 4 1,580 70 5 1.0 2,660 2,500 -160 -6 2,480 -180 -7 2.0 3,380 3,500 120 4 3,520 140 4 Control 7.0 5,120 5,100 -20 0 90 3 5,090 -30 -1 110 4 0.9 720 1,340 620 86 1,610 890 124 1.8 2,510 2,420 -90 -4 2,780 270 11 3.4 3,340 3,370 30 1 3,750 410 12 Cold 12.2 5,460 5,050 -410 -8 290 25 5,260 -200 -4 440 38 0.4 1,990 2,430 440 22 1,900 -90 -5 0.8 3,280 3,630 350 11 3,010 -270 -8 1.4 4,110 4,370 260 6 3,840 -270 -7 Hot 4.9 5,400 5,650 250 5 330 11 5,320 -80 -1 180 5 182 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.42: Error plot for the 20% F mixture, simplified method, E= 40 kJ/mol 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.43: Error plot for the 20% F mixture, simplified method, E = 25 kJ/mol 183 Table 6.28: Error using AM function based on simplified method for Type I - 0.41 mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,030 1,000 -30 -3 1,000 -30 -3 1.0 2,700 2,830 130 5 2,810 110 4 2.1 4,060 3,900 -160 -4 3,920 -140 -3 Control 7.1 4,900 4,960 60 1 100 3 4,950 50 1 80 3 0.8 540 1,060 520 96 1,490 950 176 1.8 2,830 2,860 30 1 3,240 410 14 3.4 4,040 3,820 -220 -5 4,120 80 2 Cold 12.1 5,480 4,940 -540 -10 330 28 5,030 -450 -8 470 50 0.4 2,140 2,700 560 26 1,830 -310 -14 0.8 3,670 4,120 450 12 3,570 -100 -3 1.4 4,010 4,620 610 15 4,260 250 6 Hot 5.0 4,990 5,240 250 5 470 15 5,080 90 2 190 6 184 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.44: Error plot for the Type I - 0.41 mixture, simplified method, E= 40 kJ/mol 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.45: Error plot for the Type I - 0.41 mixture, simplified method, E = 25 kJ/mol 185 Table 6.29: Error using AM function based on simplified method for Type III - 0.37 mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 4,280 4,270 -10 0 4,280 0 0 1.0 6,530 6,580 50 1 6,550 20 0 2.1 7,970 7,910 -60 -1 7,950 -20 0 Control 7.1 9,110 9,130 20 0 40 0 9,110 0 0 10 0 0.9 2,400 3,080 680 28 4,470 2,070 86 2.0 5,670 6,470 800 14 7,170 1,500 26 3.5 6,900 7,640 740 11 8,130 1,230 18 Cold 12.6 9,060 9,070 10 0 560 13 9,190 130 1 1230 33 0.3 2,860 4,520 1,660 58 2,710 -150 -5 0.8 6,400 7,940 1,540 24 7,300 900 14 1.4 6,960 8,620 1,660 24 8,220 1,260 18 Hot 5.0 8,030 9,390 1,360 17 1560 31 9,220 1,190 15 875 13 186 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.46: Error plot for the Type III - 0.37 mixture, simplified method, E= 40 kJ/mol 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure 6.47: Error plot for the Type III - 0.37 mixture, simplified method, E = 25 kJ/mol 187 Table 6.30: Comparison of AM functions for all mixtures, simplified method E = 40,000 J/mol E = 25,000 J/mol Modified ASTM Mixture ID Total AAE (psi) Total abs. EoE (%) Total AAE (psi) Total abs. EoE (%) Total AAE (psi) Total abs. EoE (%) Function that Produced Lowest Error Type I - 0.41 300 15 250 20 240 16 Modified Type I - 0.44 170 8 170 17 170 17 40,000 Type I - 0.48 260 11 200 17 200 13 25,000 20% F 230 13 240 16 200 12 Modified 30% F 210 12 240 14 210 12 40,000 or Modified 20% C 240 9 290 15 450 16 40,000 30% C 200 12 380 23 210 13 40,000 30% Slag 390 18 400 22 390 18 40,000 or Modified 50% Slag 280 16 270 21 250 17 40,000 Type III - 0.37 720 15 710 15 710 15 - Type III - 0.44 270 10 360 20 280 11 40,000 70/20/10 - 0.37 520 17 500 18 530 22 40,000 70/20/10 - 0.44 310 21 260 26 310 34 40,000 6.4 VARIABLE TEMPERATURE-SENSITIVITY MODELS As discussed in Section 2.4, researchers such as Carino (1991) and Freiesleben Hansen and Pedersen (FHP) (1977) have shown the need for two different temperature sensitivity values: one below the reference temperature and one above. In Carino?s study, he proposed using the Nurse-Saul maturity (NSM) function, with 28?F (-2.2?C) used as a datum temperature for mixtures curing below 73?F, and 55?F (12.7?C) used for 188 mixtures curing above 73?F. Carino proposed this modification to overcome the linearity of the NSM function in order to better model his data. The FHP model is used with the Arrhenius maturity function. It uses the following equations: FHP Model for Computing the Activation Energy (Previously Equation 2.9) for Tc ? 20?C (68?F) E = 33,500 J/mol, and for Tc < 20?C (68?F) E = 33,500 + 1,470 (20-Tc) J/mol. This allows for an increase in activation energy for low curing temperatures and a decreased activation energy for high curing temperatures. It was desirable to find similar models to fit the data found in this project. The least-squares regression approach was utilized for this process. The ACFs from the actual mixture data were compared to the calculated ACFs from the proposed temperature sensitivity functions. The mixtures containing silica fume were left out of the models because they do not follow the same trends as the other mixtures. The best-fit models were determined and the resulting functions are given as: Nurse-Saul Maturity Function: (Equation 6.4) for Tc ? 73?F To = 40?F, and for Tc < 73?F To = 30?F. Arrhenius Maturty Function: (Equation 6.5) for Tc ? 23?C (73?F) E = 30,300 J/mol, for Tc < 23?C (73?F) E = 30,300 + 950 (23-Tc) J/mol. 189 The functions are plotted with the ACF data from all mixtures in Figure 6.48. Along with the new models, the ACFs for the AM function using the FHP activation energy model, as well as the NSM function using 32?F (0?C) as the datum temperature, are plotted on the graph. The figure shows that both new models are essentially the same and that the FHP model produces a higher activation energy at hot temperatures than needed, except for the 30% Slag mixture, the Class C fly ash mixtures, and the 20% F mixture. It is also apparent that the datum temperature of 32?F (0?C) results in approximately the same behavior as the two best-fit models at low temperatures and, for all practical purposes, results in approximately the same behavior at high temperatures. This is essential, as a variable temperature sensitivity model complicates the maturity method and because the Nurse-Saul function is considered easier to apply than the Arrhenius maturity function. Although using a datum temperature of 32?F (0?C) with the NSM function seems to give a good average value of temperature sensitivity for curing temperatures ranging from 40?F (4?C) to 104?F (40?C), different values may be desired for maximum accuracy, if the maturity method will be used to estimate strengths in only a hot environment. For example, Figure 6.48 shows that in hot conditions, a datum temperature of 40?F (4?C) or greater may produce lower strength estimation errors for mixtures with Class F or Class C fly ash doses up to 30% than a datum temperature of 32?F (0?C). It is difficult to determine a similar trend for mixtures with GGBF slag from the data found in this study, as the 30% and 50% dose mixtures do not exhibit similar temperature sensitivity behavior. However, in hot conditions a datum temperature of 190 32?F (0?C) or less may produce less error for straight cement mixtures (Type I or Type III). 0.0 0.5 1.0 1.5 2.0 2.5 44 54 64 74 84 94 104 Temperature (?F) Age Conversion Factor E = New model E = FHP =New model = 32F Type I - 0.41 Type I - 0.44 Type I - 0.48 20% C 30% C 20% F 30% F 30% Slag 50% Slag Type III - 0.37 Type III - 0.44 70/20/10 - 0.37 70/20/10 - 0.44 Figure 6.48: Proposed temperature sensitivity functions AM (E = FHP Model) AM (E = New Model) NSM (To = 32?F [0?C]) NSM (To = New Model) To To 191 Analysis of the accuracy of the maturity method to estimate strengths using the new temperature sensitivity models is not presented here because their accuracy should be the same as or slightly better than the results presented for the NSM function using 32?F (0?C) as the datum temperature, which were presented in Section 6.3.1. However, as the new activation energy model and ACF graph show, if the Arrhenius matur ity function is used, one should use a model that allows the cold mixtures to have a higher activation energy, around 44,000 J/mol for an average curing temperature of 48?F, while allowing the hot mixtures to have a lower activation energy, in this case 30,300 J/mol. 6.5 SUMMARY AND CONCLUSIONS The maturity method, used to estimate concrete strength, was examined in four different ways using both the Nurse-Saul maturity (NSM) and the Arrhenius maturity (AM) functions. The two functions were first evaluated using current ASTM C 1074 (2004) standards to obtain mixture-specific temperature sensitivity values. Temperature sensitivity values refer to the appropriate datum temperature or activation energy, with respect to the NSM or AM functions, that are required to convert strength-age data for different curing temperatures onto one unique strength-maturity relationship. It was determined that the late-age strength loss due to high curing temperatures, referred to as crossover, resulted in a great deal of error in estimating late-age strengths (beyond 7 days of equivalent age). Thus, it was decided to only evaluate the accuracy of the maturity method based on values up to 7 days equivalent age (i.e. 5 days for the hot mixtures and 12 days for the cold). 192 This strategy led to a method referred to in this report as the Modified ASTM method. Although the datum temperatures and activation energies did not always match commonly accepted values, the errors associated with estimating strengths with the Modified ASTM method were smaller for almost all of the mixtures evaluated. This Modified ASTM method is recommended for estimating concrete strengths when a mixture-specific temperature sensitivity value is desired. The thirteen mixtures were also evaluated using a value of temperature sensitivity that was held constant for all mixtures. This evaluation was performed twice for each type of function. The NSM function was evaluated with 14?F (-10?C) and 32?F (0?C) as datum temperatures. The NSM function using the 32?F (0?C) datum temperature produced much smaller strength estimate errors than the NSM function using the 14?F (- 10?C) datum temperature. The AM function was evaluated using activation energies of 40,000 and 25,000 J/mol. The mixtures varied as to which activation energy produced the smaller error. The best use of the AM function, for most mixtures evaluated, would be to use a lower activation energy at high curing temperatures and a higher activation energy at lower curing temperatures. This approach, similar to the one proposed by Freiesleben Hansen and Pedersen (1977), led to the fourth method analyzed, which employed a variable temperature sensitivity dependent on the curing temperature. Two new temperature sensitivity models were developed to best fit the mixtures evaluated in this study. The accuracy of using the proposed temperature sensitivity functions was not given, as their behavior was very similar to using a constant datum temperature of 32?F (0?C) for the NSM function. However, if the Arrhenius maturity function is used, a variable activation energy function, like the one proposed in Equation 193 6.5, should be used. Also, it was observed that in hot conditions, mixtures with up to 30% replacement doses of Class F or Class C fly ashes may require datum temperatures around 40?F (4?C), while straight cement mixtures may require datum temperatures of 32?F (0?C) or lower. Based on the results of this study, the most accurate, practical, and straightforward method recommended for estimating concrete strength using the maturity method is the Nurse-Saul maturity function using a datum temperature of 32?F (0?C). This method works for Type I mixtures, such as ALDOT standard mixture A1-c, as well as for Type I mixtures with up to 30% replacement of Class F or C fly ash or up to 50% replacement with GGBF Slag. The Type III mixtures as well as the ternary mixtures resulted in larger errors, but these mixtures also had some of the worst crossover effects. The environmental conditions in which the concrete was mixed and cured represent extreme effects on the hydration process of the various mixtures. The results shown in this chapter show the benefits of using the maturity method to estimate concrete strength as well as the limitations. As described in the literature review, proper measures should always be taken to verify strength estimations provided by the maturity method. 194 CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS Since it?s beginnings in the 1950?s, there have been numerous studies performed to evaluate the accuracy of the maturity method to estimate concrete strengths. The idea of concrete ?maturity? began with researchers, such as Nurse (1949), Saul (1951), and McIntosh (1949). They found that given a certain well consolidated, moist-cured concrete, the strength was directly related to the length of cure and the curing temperature. This led to the Nurse-Saul function which computes a maturity index, or as it is commonly known, the Temperature-Time Factor. Later, in the 1977 Freiesleben Hansen and Pederson proposed using the Arrhenius-based maturity equation, for computing an equivalent age. Some researchers preferred the Arrhenius maturity function over the Nurse-Saul maturity function because their data showed that it more accurately modeled concrete behavior. Users of the maturity method to estimate strength on construction projects tend to favor the Nurse-Saul maturity function because of the ease of calculations. There are many limitations and complications involving both approaches to the maturity method. These issues are discussed in Section 2.3. In order to thoroughly evaluate the accuracy of the maturity method to estimate concrete strength, thirteen concrete mixtures were evaluated in this study. The mixtures were chosen in order to examine the effect of the following variables on concrete strength-gain behavior: 195 ? different cement types (Type I and III) ? water-to-cementitious materials ratios = 0.37, 0.41, 0.44, and 0.48 ? Supplementary cementing materials (SCMs) type and dosage o Class F fly ash at 20 and 30% replacement levels o Class C fly ash at 20 and 30% replacement levels o Ground-granulate blast-furnace slag at 30 and 50% replacement levels o Silica fume at a 10% replacement level in a ternary blend with 20% Class F fly ash. In the past, researchers have evaluated the maturity method based on constant curing temperatures. However, in field operations curing temperatures actually oscillate with the ambient temperature. Therefore the concrete in this study was cured under fluctuating temperatures. Three batches were made of each mixture: cold, hot, and control batches. The cold batch temperatures simulated concrete placed in winter conditions by cycling between 40?F and 55?F (4?C and 13?C) over a 24-hour period. The hot batch temperatures modeled concrete placed in summer conditions by cycling between 90?F and 105?F (32?C and 41?C) over a 24-hour period. The control batch was cured in laboratory conditions with temperatures ranging between 68?F and 73?F (20?C and 23?C). The temperature ranges were chosen in order to effectively cover the full range of practical of temperatures to be expected on Alabama Department of Transportation (ALDOT) projects. All mixing materials used for the hot and cold batches were heated or cooled to a temperature close to their respective curing environments prior to mixing. This was done to simulate conditions at a ready-mix concrete plant, where aggregates and cementitious materials are exposed to ambient 196 temperatures. Chemical admixtures were used to achieve desired slump and air content, as would be required with ready-mix concrete. All chemical admixture doses were kept constant within each mixture independent of temperature. From each batch nineteen 6x12 in. cylinders were prepared. A temperature sensor was inserted into one of the freshly prepared cylinders. The sensor was programmed to record the temperature every 30 minutes, on average. The remaining 18 cylinders were used for compressive strength testing was performed on sets of three cylinders at six different ages. The equivalent age at testing was as follows: twice the time of final set, 24 hrs, 48 hrs, 7 days, 14 days, and 28 days. These ages were adjusted for the cold and hot batches to account for the temperature dependent strength-gain. After the final compressive strength tests were completed, the temperature data were downloaded. Compressive strength test data versus chronological age was then analyzed based on the amount of late-age strength loss or gain, referred to as ?crossover,? due to early- age curing temperatures. The mixtures were also evaluated based on their apparent ?temperature sensitivity,? or the correction required to converge the data from the three batches onto a single strength-maturity relationship. The strength-age relationships for each batch were then converted to strength- maturity relationships. Maturity was computed using the Nurse-Saul and Arrhenius maturity functions. Four methods to determine temperature sensitivity values were presented. First, mixture-specific temperature sensitivity values were computed based on present ASTM C 1074 methods. These temperature sensitivity values were used to determine the unique strength-maturity (S-M) relationship for each mixture. The S-M relationship was used to estimate concrete strength of a single mixture, given the 197 temperature history of any batch. The accuracy of the maturity method was determined from the difference in the estimated strength found from the S-M relationship and the actual test strengt hs at the same maturity. The data was then reanalyzed using the ASTM methods, only slightly modified to exclude late-age data. Next, a method using commonly used constant values of temperature sensitivity for all mixtures was presented. The values chosen were based on an age conversion factor versus temperature plot for all the mixtures. The values chosen were datum temperatures of 32?F (0?C) and 14?F (-10?C) for the Nurse-Saul maturity function and activation energies of 25,000 J/mol and 40,000 J/mol for the Arrhenius maturity function. Finally, temperature dependent models used to calculate values of datum temperatures or activation energies were presented. Given the scope of this project, the maturity method provides a practical and effective means to estimate early-age concrete strengths for mixtures with various cement types, various SCM types and dosages, and various water-to-cementitious materials ratios when cured at fluctuating temperatures. 7.1 CONCLUSIONS The conclusions based on the project objectives defined in Section 1.3 are discussed in this section. The conclusions based on the effect of various fluctuating curing temperatures on concrete strength behavior are as follows: ? Each batch of all the mixtures did follow a unique strength versus age curve. ? The hot batch gained strength rapidly, but generally had lower late-age strengths when compared to those from control and cold batches. 198 ? The cold batch gained strength slowly, but continued to gain strength at late- ages. ? This phenomena is known as the ?crossover effect.? The conclusions based on the effect of various types and doses of SCMs, varying cement types, and water-to-cementitious materials ratios on the rate of strength gain at different temperatures were based on the amount of crossover and the amount of temperature sensitivity required to converge the data onto a single-strength-maturity relationship. The conclusions related on the crossover effect are as follows: ? All straight cement mixtures exhibited crossover, with long-term strength losses ranging from 7% to 12%; crossover occurred between 7 days and 16 days. ? The replacement of cement with 20% and 30% Class F fly ash for the Type I - 0.41 mixture effectively eliminated the crossover effect. ? The replacement of cement with 20% Class C fly ash for the Type I - 0.41 mixture delayed, but did not completely eliminate, the crossover effect. An increased replacement dosage of 30% Class C fly ash for the Type I - 0.41 mixture effectively eliminated crossover. ? The replacement of cement with 30% or 50% GGBF slag for the Type I - 0.41 mixture increased strength losses for the hot batch from 6% to more than 17% in some cases. ? Changing the cement from Type I to Type III for the Type I - 0.44 mixture increased the crossover effect only slightly, but greatly decreased the time at which crossover occured from 16 days to 4 days. 199 ? The replacement of cement with 20% Class F fly ash and 10% silica fume for the Type I - 0.44 mixture increased the strength loss from 7% to 31% and decreased the time at which crossover occurred from 16 days to 5 days. ? The Type III - 0.37 mixture had a 17% strength loss attributable to the crossover effect, and crossover occurred less than two days after mixing for the hot batch. ? The ternary blend prestressed concrete mixture had a long-term strength loss of 23% from the crossover effect. The conclusions observed from each mixture?s temperature sensitivity trends are as follows: ? The Type I - 0.41 and Type I - 0.48 mixtures have low temperature sensitivity for both hot and cold batches. The Type I - 0.44 had low temperature sensitivity, possibly a slightly higher sensitivity for the hot batch compared to the cold batch. ? The replacement of cement with 20% Class F fly ash for the Type I - 0.41 mixture increased the temperature sensitivity slightly. Increasing the dose to 30% Class F fly ash increased the hot batch?s temperature sensitivity. ? The replacement of cement with 20% Class C fly ash for the Type I - 0.41 mixture increased the temperature sensitivity of the cold batch, and increasing the dose to 30% Class C fly ash increased both the hot and cold batch?s temperature sensitivity significantly relative to that of the Type I - 0.41 mixture. 200 ? The replacement of cement with 30% and 50% GGBF slag for the Type I - 0.41 mixture increased the temperature sensitivity of the cold batch, but did not affect the hot batches. ? Changing the cement from Type I to Type III for the Type I - 0.44 mixture increased the temperature sensitivity for the cold batch and decreased that for the hot batch. ? The replacement of cement with 20% Class F fly ash and 10% silica fume for the Type I - 0.44 mixture increased the temperature sensitivity for the cold batch and decreased that for the hot batch. ? The Type III - 0.37 hot batch had little to no temperature sensitivity, while the cold batch required a higher temperature sensitivity. ? The ternary blend prestressed concrete mixture (70/20/10 - 0.37) had a low temperature sensitivity for the hot batch and a higher sensitivity for the cold batch. The conclusions based on the accuracy of the maturity method to estimate concrete strength for numerous mixtures with varying types and doses of SCMs, varying cement types, and water-to-cementitious materials ratios using ASTM C 1074 maturity methods are as follows: ? Average absolute percent error in estimating strengths ranged from 6% to 27% for some mixtures. ? Mixtures that produced the highest errors in estimating concrete strengths also had high strength loss due to the crossover effect. 201 It was determined that strength estimations are not accurate beyond 7 days of equivalent age. Also the late-age strength losses associated with the hot batches were believed to be affecting the temperature sensitivity determination. Thus, temperature sensitivity values were recalculated and the accuracy of strength estimations using the maturity method were reanalyzed. This was referred to as the ?Modified ASTM? method. The conclusion found from modifications to the current ASTM C 1074 procedure to handle strength estimation deficiencies the improvement in strength prediction accuracy are as follows: ? Temperature sensitivity values did change when only data up to 7 days of equivalent age was used. ? When using the Nurse-Saul maturity function, the Modified ASTM method produced less total error than the ASTM C 1074 method for all mixtures except the 20% C, Type III - 0.44, and ternary-blend mixtures. ? When using the AM function, the Modified ASTM method produced less error than the ASTM C 1074 method for all mixtures except the 20% C and ternary-blend mixtures. ? The Modified ASTM method should be used to determine mixture-specific temperature sensitivities. ? Due to the increased accuracy of the estimated strengths that resulted from using the maturity method only up to 7 days equivalent age, it was determined that strengths should not be estimated beyond 7 days of equivalent age. 202 ? Total percent errors from strength estimations ranged between 7% and 35%. However, for many mixtures the initial strength estimation for the cold batch was the main source of error. The conclusions from analysis of the accuracy of the maturity method using constant temperature sensitivity values for all mixtures are as follows: ? When using the Nurse-Saul maturity function, a datum temperature of 32?F (0?C) produced less error than that found by using 14?F (-10?C) for all mixtures. ? A datum temperature of 32?F (0?C) produced less total average error than using mixture-specific datum temperatures found from the Modified ASTM method for 8 out of the 13 mixtures evaluated. ? When using the Arrhenius maturity function, an activation energy of 40,000 J/mol for cold batches and 25,000 J/mol for hot batches produced the least error in strength estimations for 10 out of the 13 mixtures evaluated. Lastly, the use of variable temperature sensitivity models was evaluated. The conclusions from that analysis are as follows: ? The best-fit temperature sensitivity models had very similar behavior to the Nurse-Saul maturity function with a datum temperature of 32?F (0?C). ? If the Arrhenius function is used, a model should be used, such as the one proposed in Equation 6.5, which allows a higher activation energy at low temperatures and a lower activation energy at high temperatures. 203 7.2 RECOMMENDATIONS From the analysis in this study, the following recommendations were made based on the experimental results of this study: ? Strength estimations using the maturity method may not be accurate beyond 7 days of equivalent age. ? If a mixture-specific temperature sensitivity value is desired, use the Modified ASTM method, as shown in Section 6.2. ? If the Arrhenius maturity function is to be used, employ a temperature dependent model, such as the FHP model (Section 2.4.2) or the new model (Section 6.4), that allows for a high temperature sensitivity at lo w temperatures and a lower sensitivity at high temperatures. ? Select a datum temperature of 40?F (4?C) or greater when using cement- replacement dosages of 20 to 30% of Class F or Class C fly ash in hot conditions. Finally, the maturity function to be used on ALDOT projects that minimizes strength estimation errors while maintaining ease of use is the Nurse-Saul maturity function with a constant datum temperature of 32?F (0?C). 200 REFERENCES AASHTO M 43. 2003. Sizes of Aggregate for Road and Bridge Construction. Standard Specifications for Transportation Materials and Methods of Sampling and Testing, Part I A: Specifications, 23rd Edition. Washington, D.C. AASHTO T 22. 2003. Compressive Strength of Cylindrical Concrete Specimens. In Standard Specifications for Transportation Materials and Methods of Sampling and Testing, Part 2 A: Tests, 23rd Edition. Washington, D.C. ALDOT Section 501.02. 2002. Structural Portland Cement Concrete. Standard Specifications for Highway Construction. Montgomery, Alabama. Alexander, K.M., and J.H. Taplin. 1962. Concrete Strength, Paste Strength, Cement Hydration, and the Maturity Rule. Australian Journal of Applied Science, 13: 277-284. ASTM C 33. 2004. Standard Specification for Concrete Aggregates. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 39. 2004. Standard Test Methods for Compressive Strength of Cylindrical Concrete Specimens. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 127. 2004. Standard Test Method for Density, Relative Density (Specific Gravity), and Absorption of Coarse Aggregate. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 128. 2004. Standard Test Method for Density, Relative Density (Specific Gravity), and Absorption of Fine Aggregate. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 138. 2004. Standard Test Method for Density (Unit Weight), Yield, and Air Content (Gravimetric) of Concrete. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 143. 2004. Standard Test Method for Slump of Hydraulic-Cement Concrete. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. 201 ASTM C 192. 2004. Standard Practice for Making and Curing Concrete Test Specimens in the Laboratory. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 231. 2004. Standard Test Method for Air Content of Freshly Mixed Concrete by the Pressure Method. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 684. 2004. Standard Test Method for Making, Accelerated Curing, and Testing Concrete Compression Test Specimens. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 803. 2004. Standard Test Method for Penetration Resistance of Hardened Concrete. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 873. 2004. Standard Test Method for Compressive Strength of Concrete Cylinders Cast in Place in Cylindrical Molds. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 900. 2004. Standard Test Method for Pullout Strength of Hardened Concrete. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 918. 2004. Standard Test Method for Measuring Early-Age Compressive Strength and Projecting Later-Age Strength. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 1064. 2004. Standard Test Method for Temperature of Freshly Mixed Hydraulic-Cement Concrete. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 1074. 2004. Standard Practice for Estimating Concrete Strength by the Maturity Method. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. ASTM C 1231. 2004. Standard Practice for Use of Unbonded Caps in Determination of Compressive Strength of Hardened Concrete Cylinders. Annual Book of ASTM Standards, West Conshohocken, Pennsylvania. Bergstrom, S.G. 1953. Curing Temperature, Age and Strength of Concrete. Magazine of Concrete Research, Dec. pp. 61-66. Carino, N.J. 1984. Maturity Method: Theory and Application. Cement, Concrete and Aggregates, 6 (2): 61-73. Carino, N.J. 1991. The Maturity Method. In Handbook on Nondestructive Testing of Concrete, eds. N.J. Carino and V.M. Malhotra, CRC Press: Boca Raton, Florida. pp. 101-146. 202 Carino, N.J. 1997. Nondestructive Test Methods. In Concrete Construction Engineering Handbook, ed. E.G. Nawy, CRC Press: Boca Raton, FL. pp. 16-24. Carino, N.J., and H.S. Lew. 1983. Temperature Effects on Strength-Maturity Relations of Mortar. ACI Journal, 3: 177-182. Carino, N.J., and R.C. Tank. 1992. Maturity Functions for Concretes Made with Various Cements and Admixtures. ACI Materials Journal. 89 (2): 188-196. Cordon, W.A. 1946. Entrained Air ?A Factor in the Design of Concrete Mixes. Materials Laboratories Report No. C-310, Research and Geology Division, Bureau of Reclamation. Denver. Freiesleben Hansen, P., and E.J. Pedersen. 1977. Maturity Computer for Controlled Curing and Hardening of Concrete. Journal of the Nordic Concrete Federation, No. 1: 21-25. Gonnerman, H.F. and E.C. Shuman. 1928. Flexure and Tension Tests of Plain Concrete. Report of the Director of Research, Skokie, Illinois: Portland Cement Association. Nov., pp. 149-163. Guo, C. 1989. Maturity of Concrete: Method for Predicting Early-Stage Strength. ACI Materials Journal, 86 (4): 341-353. Holland, T.C. 2005. Silica Fume User?s Manual. Federal Highway Administration / Silica Fume Association. Washington, D.C. pp. 63-65. Iowa Department of Transportation. 2000. Method of Testing the Strength of Portland Cement Concrete Using the Maturity Method. Materials Instructional Memorandums, 383. 10 pp. Jonasson, J.E. 1985. Early Strength Growth in Concrete?Preliminary Test Results Concerning Hardening at Elevated Temperatures. Proceeding of the 3rd International RILEM Symposium on Winter Concreting, Espoo, Finland. pp. 249- 254. Kjellsen, K.O., and R.J. Detwiler. 1992. Reaction Kinetics of Portland Cement Mortars Hydrated at Different Temperatures. Cement and Concrete Research, 22 (1): 112-120. Kjellsen, K.O., and R.J. Detwiler. 1993. Later-Age Strength Prediction by a Modified Maturity Model. ACI Materials Journal, 90 (3): 220-227. Kjellsen, K.O., R.J. Detwiler, O.E. Gj?rv. 1991. Development of Microstructures in Plain Cement Pastes Hydrated at Different Temperatures. Cement and Concrete Research, 21: 179-189. 203 Klieger, P. 1958. Effect of Mixing and Curing Temperature on Concrete Strength. Journal of the American Concrete Institute, 29 (12): 1063-1081. Kosmatka, S.A., B. Kerkhoff, and W.C. Panarese. 2002. Design and Control of Concrete Mixtures, 14 ed. Portland Cement Association: Skokie, Illinois. McIntosh, J.D. 1949. Electrical Curing of Concrete. Magazine of Concrete Research, Jan. pp. 21-28. McIntosh, J.D. 1956. Effect of Low-Temperature Curing on the Compressive Strength of Concrete. Proceedings of the RILEM Symposium: Winter Concreting, Copenhagen, Session BII, 17 pp. Mindess, S., J.F. Young, , and D. Darwin. 2002. Concrete. Prentice-Hall Inc: New Jersey, 671 pp. Nurse, R.W. 1949. Steam Curing of Concrete. Magazine of Concrete Research, June pp. 79-88. Nykanen, A. 1956. Hardening of Concrete at Different Temperatures, Especially Below the Freezing Point. Proceedings of the RILEM Symposium: Winter Concreting, Copenhagen, Session BII. Plowman, J.M. 1956. Maturity and the Strength of Concrete. Magazine of Concrete Research, Mar. pp.13-22. Rastrop, E. 1954. Heat of Hydration in Concrete. Magazine of Concrete Research, 6(17): p 79. Springenschmid, R. ed.1998. Prevention of Thermal Cracking in Concrete at Early Ages. New York: Routledge. RILEM, Report 15. pp. 324-325. Saul, A.G.A. 1951. Principles Underlying the Steam Curing of Concrete at Atmospheric Pressure. Magazine of Concrete Research, Mar. pp. 127-140. Tank, R.C., and N.J. Carino. 1991. Rate Constant Functions for Strength Development of Concrete. ACI Materials Journal, 88 (1): 74-83. Texas Department of Transportation, Materials and Tests Division. 1999. Estimating Concrete Strength by the Maturity Method, Test Method Tex 426-A. Manual of Test Procedures, Vol. II. Top?u, I.B., and A. Ugurlu. 2003. Effect of the use of Mineral Filler on the Properties of Concrete. Cement and Concrete Research, 33: 1071-1075. Verbeck, G.J., and R.H. Helmuth. 1968. Structures and Physical Properties of Cement Paste. Proceedings of the Fifth International Symposium on the Chemistry of Cement, Tokyo. pp. 1-32. 204 APPENDICES 205 APPENDIX A: COMPRESSIVE STRENGTH TEST RESULTS Appendix A-1: Compressive strength test results for Type I- 0.41 mixture Batch ID Age (days) Compressive Strength (psi) 0.4 2,076 2,123 2,225 0.8 3,653 3,721 3,646 1.4 4,051 4,071 3,935 5.0 5,119 4,944 4,928 10.2 5,557 5,330 5,353 Hot 20.1 5,353 5,836 5,684 0.5 1,010 1,040 1,040 1.0 2,650 2,780 2,690 2.1 3,990 4,210 3,990 7.1 4,840 4,990 4,900 14.0 5,450 5,850 5,770 Control 28.1 6,220 6,130 6,160 0.8 550 850* 540 1.8 2,770 2,860 2,880 3.4 3,360* 4,170 3,930 12.1 5,440 5,520 5,050* 24.8 6,150 5,740 6,030 Cold 49.0 6,160* 6,600 6,650 *Not used for average strength Appendix A-2: Compressive strength test results for Type I- 0.44 mixture Batch ID Age (days) Compressive Strength (psi) 0.3 1,250 1,280 1,300 0.9 2,610 2,590 2,520 1.5 3,450 3,380 3,340 5.0 4,380 4,380 4,400 12.1 5,070 5,100 5,100 Hot 20.1 5,410 5, 400 5,380 0.5 850 810 1,090* 1.0 1,970 1,930 2,090 2.0 2,890 2,790 2,470* 7.3 4,350 4,080 4,220 14.2 5,180 4,820 4,930 Control 28.0 5,780 5,760 - 0.9 410 440 540* 1.9 2,100 1,780 2,010 3.5 3,180 2,950 3,140 12.4 4,600 4,560 5,030 25.0 5,730 5,490 5,800 Cold 49.2 6,210 6,470 6,560 *Not used for average strength 206 Appendix A-3: Compressive strength test results for Type I- 0.48 mixture Batch ID Age (days) Compressive Strength (psi) 0.3 780* 870 850 0.8 2,170 2,260 2,120 1.4 2,730 2,740 2,880 5.1 4,020 4,150 3,880 9.8 4,55760 4,770 4,710 Hot 19.8 5,050 5,200 5,000 0.5 630 640 570* 1.0 1,690 1,560* 1,740 2.0 2,650 2,510 2,550 7.4 4,280 4,110 3,990 14.2 4,710 4,790 4,600 Control 28.3 5,930 5,640 5,580 0.9 420 550* 420 1.9 1,810* 1,570 1,640 3.6 3,110 3,120 2,900 12.0 4,450 4,780 4,630 25.1 6,160 6,010 5,510* Cold 49.1 7,030 7,310 7,300 *Not used for average strength Appendix A-4: Compressive strength test results for 20% F mixture Batch ID Age (days) Compressive Strength (psi) 0.4 1,950 2,110 1,920 0.8 3,490 3,180 3,190 1.4 4,020 4,200 4,120 4.9 5,400 5,560 5,260 10.0 6,240 6,290 6,020 Hot 19.8 7,150 6,990 7,070 0.6 1,510 1,280* 1,530 1.0 2,670 2,370* 2,650 2.0 3,310 3,560 3,300 7.0 5,280 4,900 5,200 14.1 5,980 6,070 5,750 Control 28.2 6,830 7,260 6,810 0.9 730 730 480* 1.8 2,510 2,210* 2,530 3.4 3,420 3,340 3,270 12.2 5,670 5,430 5,300 24.9 6,030 6,180 6,110 Cold 49.0 6,600 6,690 6,400 *Not used for average strength 207 Appendix A-5: Compressive strength test results for 30% F mixture Batch ID Age (days) Compressive Strength (psi) 0.4 1,370 1,350 1,350 0.8 2,770 2,860 2,960 1.5 3,690 3,600 3,600 5.0 4,970 5,050 4,870 10.1 6,240 5,830* 6,140 Hot 20.0 7,470* 7,070 6,860 0.6 1,180 1,360* 1,190 1.1 2,240 2,170 2,200 2.0 2,820 3,250* 2,910 7.0 4,420 4,570 4,970* 14.1 5,430 5,360 5,490 Control 28.2 6,550 6,510 6,390 0.9 460* 690 670 1.9 2,250* 1,940 2,050 3.5 3,163 3,122 3,000 12.2 5,040 5,140 4,800 25.0 5,500 5,470 5,470 Cold 49.0 6,270 6,320 6,470 *Not used for average strength Appendix A-6: Compressive strength test results for 20% C mixture Batch ID Age (days) Compressive Strength (psi) 0.5 1,880 2,020 1,910 0.8 3,210* 3,630 3,590 1.4 3,820 3,980 4,370* 5.0 4,990* 5,460 5,280 10.3 5,850 5,910 5,800 Hot 20.0 6,510 6,280 6,530 0.6 880* 1,180 1,260 1.0 2,410* 2,200 2,230 2.0 3,740 3,780 3,700 7.0 5,420 5,220 5,350 14.1 5,740 6,070 5,900 Control 28.0 6,560 6,800 6,530 1.2 1,050 1,040 670* 1.8 2,170* 1,710 1,830 3.4 3,240 2,760* 3,040 12.0 5,450 5,360 5,450 25.9 6,050 5,870 5,690 Cold 49.0 6,010* 6,350 6,580 *Not used for average strength 208 Appendix A-7: Compressive strength test results for 30% C mixture Batch ID Age (days) Compressive Strength (psi) 0.4 1,030 1,040 1,100 0.8 2,720 2,690 2,740 1.4 4,520* 3,930 3,880 5.0 5,991 5,940 5,700 9.9 5,940 5,990 6,300 Hot 19.9 7,540 6,690* 7,490 0.6 680 700 660 1.0 1,620* 1,990 1,900 2.0 3,090 3,040 3,470* 7.2 5,070 5,450 5,240 14.1 5,690 5,950 6,380* Control 28.1 6,990 5,940* 6,960 1.3 560 840* 590 1.9 1,710 1,300* 1,660 3.5 2,540 2,700 2,630 12.1 4,690 5,140* 4,630 25.0 6,360 5,710* 6,360 Cold 49.2 6,730 6,590 6,540 *Not used for average strength Appendix A-8: Compressive strength test results for 30% Slag mixture Batch ID Age (days) Compressive Strength (psi) 0.4 1,540 1,640* 1,470 0.8 3,100 2,830* 3,170 2.9 4,840 5,090 5,060 4.9 5,120 4,840* 5,360 10.8 5,840 6,040 6,010 Hot 19.8 6,090 6,050 5,900* 0.6 1,420 1,440 1,270* 1.0 2,520* 2,700 2,760 2.0 3,620 3,670 3,680 7.1 5,610 5,810 5,930 14.1 6,340 6,560 6,950* Control 28.0 7,050 7,170 7,340 0.9 630 610 590 1.9 2,040 2,040 2,090 3.6 3,380 3,370 3,450 12.2 5,100 4,990 5,170 25.1 6,460 6,730 6,740 Cold 49.4 7,710 7,810 7,750 *Not used for average strength 209 Appendix A-9: Compressive strength test results for 50% Slag mixture Batch ID Age (days) Compressive Strength (psi) 0.3 620 600 620 0.8 2,290 2,390 2,320 1.4 3,200 3,220 3,320 5.0 5,920 6,210 5,610* 10.1 6,340 6,720 6,430 Hot 19.8 6,510 6,630 6,510 0.5 740 750 790 1.0 1,710 1,610 1,730 2.1 3,020 3,110 3,080 7.2 5,740 5,850 5,750 14.2 7,160 7,000 7,130 Control 28.1 7,660 7,860 8,10 0.9 340 360* 330 1.9 1,400 1,330 1,300 3.4 2,740* 2,420 2,490 12.3 5,070 4,860 4,980 24.9 6,720 6,820 6,700 Cold 49.0 8,510 8,680 8,510 *Not used for average strength Appendix A-10: Compressive strength test results for Type III - 0.37 mixture Batch ID Age (days) Compressive Strength (psi) 0.3 2,820 2,920 3,400* 0.8 6,400 6,300 6,500 1.4 6,980 6,940 6,990 5.0 8,050 8,110 7,940 10.0 8,930 8,520 8,400 Hot 19.8 9,190 8,900 9,040 0.6 4,230 4,970* 4,340 1.0 6,280 6,620 6,710 2.1 8,150 7,880 7,890 7.1 9,360 9,060 8,910 14.3 10,070 9,760 10,470 Control 28.2 10,690 11,510 11,180 0.9 2,360 2,450 1,500* 2.0 5,690 5,660 4,500* 3.5 6,850 6,950 5,790* 12.6 9,590 8,540 7,370* 25.1 10,680 9,880 7,820* Cold 49.4 10,560 10,490 9,180* *Not used for average strength 210 Appendix A-11: Compressive strength test results for Type III - 0.44 mixture Batch ID Age (days) Compressive Strength (psi) 0.3 1,920* 2,360 2,440 0.7 3,340 3,120 3,190 1.4 3,650 3,830 4,040* 5.3 5,170 5,010 5,140 10.0 5,220 5,850* 5,370 Hot 20.2 5,560 5,480 5,480 0.4 900* 720 750 1.0 3,070 2,920 3,150 2.0 3,660 3,850 3,980 7.2 4,960 5,210 5,150 14.0 5,990 5,760 5,590 Control 28.2 6,120 6,560 6,070 1.0 720 970* 780 1.8 2,310 2,330 2,490 3.4 3,670 3,540 3,770 12.1 5,780 5,400 5,730 24.9 6,410 6,390 6,500 Cold 49.2 7,220 6,870 7,140 *Not used for average strength Appendix A-12: Compressive strength test results for 70/20/10 - 0.37 mixture Batch ID Age (days) Compressive Strength (psi) 0.3 1,710 1,570 1,910* 0.7 3,970* 3,460 3,620 1.4 4,960 4,900 5,220 4.9 7,590 7,320 7,280 9.8 8,030 8,370 8,330 Hot 20.0 8,670 8,270 8,550 0.5 1,270 1,350 830* 1.0 3,630 3,680 3,340* 2.1 5,190 5,360 5,190 7.0 7,580 7,760 7,560* 14.1 10,040 10,300 10,020 Control 28.1 10,760 10,840 11,280* 0.9 1,070 1,030 1,170* 1.8 3,070 3,150 2,980 3.5 4,450 4,740 4,780 12.2 6,980 7,040 7,750* 25.0 8,920 9,070 8,630 Cold 49.2 11,020 10,630 10,820 *Not used for average strength 211 Appendix A-13: Compressive strength test results for 70/20/10 - 0.44 mixture Batch ID Age (days) Compressive Strength (psi) 0.3 910* 960 1,010 0.8 1,870 1,890 1,850 1.5 2,430* 2,830 2,630 5.0 4,510 4,510 4,510 10.2 4,850 4,880 4,820 Hot 19.9 5,320 5,360 5,240 0.4 500 530 530 1.0 1,820 1,810 1,700 2.0 2,540 2,750 2,550 7.2 4,610 4,730 4,710 13.9 6,340 6,140 6,290 Control 27.9 7,070 7,280 7,090 1.0 440* 380 380 1.8 1,300* 1,470 1,440 3.4 2,490 2,640 2,340* 12.1 4,410 4,530 4,600 24.9 5,730 5,950 5,800 Cold 48.9 7,360 7,580 7,150 *Not used for average strength 212 APPENDIX B: STRENGTH-MATURITY RELATIONSHIPS FOR NURSE-SAUL MATURITY FUNCTION, ASTM C 1074 METHOD 213 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship 73 ?F Cold Hot (b) Figure B-1: Type I - 0.44 mixture, (a) compressive strength vs. age, (b) strength-maturity plot, ASTM C 1074 method To = 35?F 214 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-2: Type I - 0.48 mixture, (a) compressive strength vs. age, (b) strength-maturity plot, ASTM C 1074 method To = 39?F 215 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-3: 30% F mixture, (a) compressive strength vs. age, (b) strength-maturity plot, ASTM C 1074 method To = 19?F 216 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-4: 20% C mixture, (a) compressive strength vs. age, (b) strength-maturity plot, ASTM C 1074 method To = 37?F 217 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-5: 30% C mixture, (a) compressive strength vs. age, (b) strength-maturity plot, ASTM C 1074 method To = 42?F 218 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-6: 30% Slag mixture, (a) compressive strength vs. age, (b) strength-maturity plot, ASTM C 1074 method To = 46?F 219 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-7: 50% Slag mixture, (a) compressive strength vs. age, (b) strength-maturity plot, ASTM C 1074 method To = 47?F 220 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-8: Type III - 0.44 mixture, (a) compressive strength vs. age, (b) strength- maturity plot, ASTM C 1074 method To = 37?F 221 0 2,000 4,000 6,000 8,000 10,000 12,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 2,000 4,000 6,000 8,000 10,000 12,000 0 5,000 10,000 15,000 20,000 25,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-9: 70/20/10 - 0.37 mixture, (a) compressive strength vs. age, (b) strength- maturity plot, ASTM C 1074 method To = 39?F 222 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Concrete Age (hours) Compressive Strength (psi) Regression Lines Control Data Cold Data Hot Data (a) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data (b) Figure B-10: 70/20/10 - 0.44 mixture, (a) compressive strength vs. age, (b) strength- maturity plot, ASTM C 1074 method To = 43?F 223 APPENDIX C: STRENGTH-MATURITY RELATIONSHIPS FOR ARRHENIUS MATURITY FUNCTION, ASTM C 1074 METHOD 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 200 400 600 800 1,000 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-1: Strength-maturity plot, ASTM C 1074 method for Type I - 0.44 mixture E = 34,800 J/mol 224 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-2: Strength-maturity plot, ASTM C 1074 method for Type I - 0.48 mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-3: Strength-maturity plot, ASTM C 1074 method for 30% F mixture E = 42,300 J/mol E = 23,200 J/mol 225 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 200 400 600 800 1,000 1,200 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-4: Strength-maturity plot, ASTM C 1074 method for 20% C mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 1,200 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-5: Strength-maturity plot, ASTM C 1074 method for 30% C mixture E = 37,800 J/mol E = 45,100 J/mol 226 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 200 400 600 800 1,000 1,200 1,400 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-6: Strength-maturity plot, ASTM C 1074 method for 30% Slag mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 300 600 900 1,200 1,500 1,800 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-7: Strength-maturity plot, ASTM C 1074 method for 50% Slag mixture E = 55,700 J/mol E = 61,500 J/mol 227 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 200 400 600 800 1,000 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-8: Strength-maturity plot, ASTM C 1074 method for Type III - 0.44 mixture 0 2,000 4,000 6,000 8,000 10,000 12,000 0 200 400 600 800 1,000 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-9: Strength-maturity plot, ASTM C 1074 method for 70/20/10 - 0.37 mixture E = 42,900 J/mol E = 42,600 J/mol 228 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 200 400 600 800 1,000 1,200 1,400 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure C-10: Strength-maturity plot, ASTM C 1074 method for 70/20/10 - 0.44 mixture E = 47,300 J/mol 229 APPENDIX D: ERROR TABLES FOR ASTM C 1074 METHOD Table D-1: Error using NSM function based on ASTM C 1074 methods for the Type I - 0.44 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 830 990 160 19 1.0 1990 1800 -190 -10 2.0 2840 2710 -130 -5 7.3 4210 4470 260 6 14.2 4970 5110 140 3 Control 28.0 5760 5520 -240 -4 190 8 0.9 420 800 380 90 1.9 2050 1660 -390 -19 3.5 3090 2410 -680 -22 12.4 4580 3980 -600 -13 25.0 5670 4760 -910 -16 Cold 49.2 6410 5360 -1050 -16 670 30 0.3 1270 930 -340 -27 0.9 2570 2300 -270 -11 1.5 3390 3160 -230 -7 5.0 4380 4680 300 7 12.1 5090 5390 300 6 Hot 20.1 5390 5630 240 4 280 10 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-1: Estimated strength versus actual strength for Type I - 0.44 mixture 230 Table D-2: Error using NSM function based on ASTM C 1074 methods for the Type I - 0.48 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 630 780 150 24 1.0 1710 1540 -170 -10 2.0 2570 2440 -130 -5 7.4 4120 4260 140 3 14.2 4700 4960 260 6 Control 28.3 5710 5450 -260 -5 190 9 0.9 410 580 170 41 1.9 1600 1210 -390 -24 3.6 3040 1770 -1270 -42 12.0 4610 3170 -1440 -31 25.1 6080 4180 -1900 -31 Cold 49.1 7210 4900 -2310 -32 1247 34 0.3 850 780 -70 -8 0.8 2180 1980 -200 -9 1.4 2770 2740 -30 -1 5.1 4010 4530 520 13 9.8 4670 5160 490 10 Hot 19.8 5080 5580 500 10 302 9 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-2: Estimated strength versus actual strength for Type I - 0.48 mixture 231 Table D-3: Error using NSM function based on ASTM C 1074 methods for the 30% F mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1180 1220 40 3 1.1 2200 1910 -290 -13 2.0 2860 2810 -50 -2 7.0 4490 4810 320 7 14.1 5420 5620 200 4 Control 28.2 6480 6160 -320 -5 200 6 0.9 670 1250 580 87 1.9 1990 2170 180 9 3.5 3090 3010 -80 -3 12.2 4990 4860 -130 -3 25.0 5480 5840 360 7 Cold 49.0 6350 6250 -100 -2 240 18 0.4 1350 1220 -130 -10 0.8 2860 2180 -680 -24 1.5 3620 3080 -540 -15 5.0 4960 5000 40 1 10.1 6190 5780 -410 -7 Hot 20.0 6960 6250 -710 -10 420 11 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-3: Estimated strength versus actual strength for 30% F mixture 232 Table D-4: Error using NSM function based on ASTM C 1074 methods for the 20% C mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1220 1200 -20 -2 1.0 2210 2320 110 5 2.0 3730 3540 -190 -5 7.0 5330 5440 110 2 14.1 5900 6060 160 3 Control 28.0 6620 6440 -180 -3 130 3 1.2 1040 1010 -30 -3 1.8 1760 1810 50 3 3.4 3140 2610 -530 -17 12.0 5410 4590 -820 -15 25.9 5870 5560 -310 -5 Cold 49.0 6460 6330 -130 -2 310 8 0.5 1930 1780 -150 -8 0.8 3600 3040 -560 -16 1.4 3900 4030 130 3 5.0 5360 5740 380 7 10.3 5850 6310 460 8 Hot 20.0 6440 6570 130 2 300 7 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-4: Estimated strength versus actual strength for 20% C mixture 233 Table D-5: Error using NSM function based on ASTM C 1074 methods for the 30% C mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 670 790 120 18 1.0 1940 1790 -150 -8 2.0 3060 3020 -40 -1 7.2 5250 5260 10 0 14.1 5820 6160 340 6 Control 28.1 6970 6690 -280 -4 160 6 1.3 570 330 -240 -42 1.9 1680 920 -760 -45 3.5 2620 1750 -870 -33 12.1 4650 4460 -190 -4 25.0 6350 6100 -250 -4 Cold 49.2 6610 6380 -230 -3 420 22 0.4 1050 1140 90 9 0.8 2710 2550 -160 -6 1.4 3900 3700 -200 -5 5.0 5870 5880 10 0 9.9 6070 6520 450 7 Hot 19.9 7510 6890 -620 -8 260 6 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot 10% Error 20% Error Figure D-5: Estimated strength versus actual strength for 30% C mixture 234 Table D-6: Error using NSM function based on ASTM C 1074 methods for the 30% Slag mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1430 1500 70 5 1.0 2730 2580 -150 -5 2.0 3650 3700 50 1 7.1 5770 5780 10 0 14.1 6450 6560 110 2 Control 28.0 7180 7090 -90 -1 80 2 0.9 610 640 30 5 1.9 2050 1450 -600 -29 3.6 3400 2040 -1360 -40 12.2 5080 3590 -1490 -29 25.1 6640 4700 -1940 -29 Cold 49.4 7750 5750 -2000 -26 1240 26 0.4 1500 1760 260 17 0.8 3130 3260 130 4 2.9 4990 5640 650 13 4.9 5240 6300 1060 20 10.8 5960 6990 1030 17 Hot 19.8 6060 7300 1240 20 730 15 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-6: Estimated strength versus actual strength for 30% Slag mixture 235 Table D-7: Error using NSM function based on ASTM C 1074 methods for the 50% Slag mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 760 740 -20 -3 1.0 1680 1690 10 1 2.1 3060 3070 10 0 7.2 5770 5790 20 0 14.2 7090 7020 -70 -1 Control 28.1 7840 7880 40 1 30 1 0.9 330 0 -330 -100 1.9 1340 280 -1060 -79 3.4 2450 740 -1710 -70 12.3 4960 2010 -2950 -59 24.9 6740 3270 -3470 -51 Cold 49.0 8560 6880 -1680 -20 1870 63 0.3 610 770 160 26 0.8 2330 2660 330 14 1.4 3240 3900 660 20 5.0 6060 6710 650 11 10.1 6490 7730 1240 19 Hot 19.8 6540 8350 1810 28 810 20 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-7: Estimated strength versus actual strength for 50% Slag mixture 236 Table D-8: Error using NSM function based on ASTM C 1074 methods for the Type III - 0.44 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 730 800 70 10 1.0 3040 2820 -220 -7 2.0 3820 3920 100 3 7.2 5100 5370 270 5 14.0 5780 5790 10 0 Control 28.2 6240 6030 -210 -3 150 5 1.0 750 980 230 31 1.8 2370 2230 -140 -6 3.4 3660 3090 -570 -16 12.1 5630 4800 -830 -15 24.9 6430 5450 -980 -15 Cold 49.2 7070 5840 -1230 -17 660 17 0.3 2400 1300 -1100 -46 0.7 3210 3020 -190 -6 1.4 3740 4110 370 10 5.3 5100 5560 460 9 10.0 5290 5890 600 11 Hot 20.2 5500 6040 540 10 540 15 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-8: Estimated strength versus actual strength for Type III - 0.44 mixture 237 Table D-9: Error using NSM function based on ASTM C 1074 methods for the 70/20/10 - 0.37 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1300 1550 250 19 1.0 3650 3320 -330 -9 2.1 5240 5050 -190 -4 7.0 7670 8320 650 8 14.1 10110 9780 -330 -3 Control 28.1 10800 10750 -50 0 300 7 0.9 1040 1520 480 46 1.8 3060 2810 -250 -8 3.5 4650 3790 -860 -18 12.2 7000 6700 -300 -4 25.0 8870 8500 -370 -4 Cold 49.2 10820 9850 -970 -9 538 15 0.3 1640 1650 10 1 0.7 3540 4010 470 13 1.4 5020 5820 800 16 4.9 7390 9090 1700 23 9.8 8240 10350 2110 26 Hot 20.0 8490 10930 2440 29 1255 18 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-9: Estimated strength versus actual strength for 70/20/10 - 0.37 mixture 238 Table D-10: Error using NSM function based on ASTM C 1074 methods for the 70/20/10 - 0.44 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 510 650 140 27 1.0 1770 1600 -170 -10 2.0 2610 2540 -70 -3 7.2 4680 4910 230 5 13.9 6250 6120 -130 -2 Control 27.9 7140 7140 0 0 120 8 1.0 370 620 250 68 1.8 1450 1070 -380 -26 3.4 2560 1500 -1060 -41 12.1 4510 3150 -1360 -30 24.9 5820 4420 -1400 -24 Cold 48.9 7360 5680 -1680 -23 1020 35 0.3 980 920 -60 -6 0.8 1860 2050 190 10 1.5 2720 3230 510 19 5.0 4500 5700 1200 27 10.2 4840 6890 2050 42 Hot 19.9 5300 7620 2320 44 1060 25 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-10: Estimated strength versus actual strength for 70/20/10 - 0.44 mixture 239 Table D-11: Error using AM function based on ASTM C 1074 methods for the Type I - 0.44 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 830 980 150 18 1.0 1990 1800 -190 -10 2.0 2840 2710 -130 -5 7.3 4210 4460 250 6 14.2 4970 5110 140 3 Control 28.0 5760 5530 -230 -4 180 7 0.9 420 900 480 114 1.9 2050 1830 -220 -11 3.5 3090 2690 -400 -13 12.4 4580 4380 -200 -4 25.0 5670 5080 -590 -10 Cold 49.2 6410 5530 -880 -14 460 28 0.3 1270 1150 -120 -9 0.9 2570 2610 40 2 1.5 3390 3450 60 2 5.0 4380 4830 450 10 12.1 5090 5470 380 7 Hot 20.1 5390 5680 290 5 220 6 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-11: Estimated strength versus actual strength for Type I - 0.44 mixture 240 Table D-12: Error using AM function based on ASTM C 1074 methods for the Type I - 0.48 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 630 780 150 24 1.0 1710 1550 -160 -9 2.0 2570 2440 -130 -5 7.4 4120 4250 130 3 14.2 4700 4960 260 6 Control 28.3 5710 5450 -260 -5 1820 9 0.9 410 670 260 63 1.9 1600 1380 -220 -14 3.6 3040 2110 -930 -31 12.0 4610 3810 -800 -17 25.1 6080 4730 -1350 -22 Cold 49.1 7210 5300 -1910 -26 910 29 0.3 850 1000 150 18 0.8 2180 2450 270 12 1.4 2770 3180 410 15 5.1 4010 4810 800 20 9.8 4670 5340 670 14 Hot 19.8 5080 5690 610 12 490 15 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-12: Estimated strength versus actual strength for Type I - 0.48 mixture 241 Table D-13: Error using AM function based on ASTM C 1074 methods for the 30% F mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1180 1220 40 3 1.1 2200 1910 -290 -13 2.0 2860 2810 -50 -2 7.0 4490 4810 320 7 14.1 5420 5620 200 4 Control 28.2 6480 6160 -320 -5 200 6 0.9 670 1320 650 97 1.9 1990 2280 290 15 3.5 3090 3170 80 3 12.2 4990 5040 50 1 25.0 5480 5900 420 8 Cold 49.0 6350 6300 -50 -1 260 21 0.4 1350 1290 -60 -4 0.8 2860 2280 -580 -20 1.5 3620 3190 -430 -12 5.0 4960 5080 120 2 10.1 6190 5820 -370 -6 Hot 20.0 6960 6280 -680 -10 370 9 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-13: Estimated strength versus actual strength for 30% F mixture 242 Table D-14: Error using AM function based on ASTM C 1074 methods for the 20% C mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1220 1200 -20 -2 1.0 2210 2330 120 5 2.0 3730 3540 -190 -5 7.0 5330 5440 110 2 14.1 5900 6060 160 3 Control 28.0 6620 6440 -180 -3 130 3 1.2 1040 1310 270 26 1.8 1760 2140 380 22 3.4 3140 3230 90 3 12.0 5410 5240 -170 -3 25.9 5870 6010 140 2 Cold 49.0 6460 6460 0 0 180 9 0.5 1930 2210 280 15 0.8 3600 3530 -70 -2 1.4 3900 4420 520 13 5.0 5360 5910 550 10 10.3 5850 6420 570 10 Hot 20.0 6440 6630 190 3 360 9 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-14: Estimated strength versus actual strength for 20% C mixture 243 Table D-15: Error using AM function based on ASTM C 1074 methods for the 30% C mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 670 780 110 16 1.0 1940 1800 -140 -7 2.0 3060 3020 -40 -1 7.2 5250 5260 10 0 14.1 5820 6150 330 6 Control 28.1 6970 6690 -280 -4 150 6 1.3 570 740 170 30 1.9 1680 1480 -200 -12 3.5 2620 2620 0 0 12.1 4650 5160 510 11 25.0 6350 6310 -40 -1 Cold 49.2 6610 6680 70 1 170 9 0.4 1050 1610 560 53 0.8 2710 3250 540 20 1.4 3900 4300 400 10 5.0 5870 6200 330 6 9.9 6070 6700 630 10 Hot 19.9 7510 6990 -520 -7 500 18 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-15: Estimated strength versus actual strength for 30% C mixture 244 Table D-16: Error using AM function based on ASTM C 1074 methods for the 30% Slag mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1430 1490 60 4 1.0 2730 2610 -120 -4 2.0 3650 3700 50 1 7.1 5770 5770 0 0 14.1 6450 6560 110 2 Control 28.0 7180 7100 -80 -1 70 2 0.9 610 840 230 38 1.9 2050 1900 -150 -7 3.6 3400 2910 -490 -14 12.2 5080 5100 20 0 25.1 6640 6150 -490 -7 Cold 49.4 7750 6830 -920 -12 380 13 0.4 1500 2660 1160 77 0.8 3130 4250 1120 36 2.9 4990 6260 1270 25 4.9 5240 6730 1490 28 10.8 5960 7230 1270 21 Hot 19.8 6060 7450 1390 23 1280 25 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-16: Estimated strength versus actual strength for 30% Slag mixture 245 Table D-17: Error using AM function based on ASTM C 1074 methods for the 50% Slag mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 760 730 -30 -4 1.0 1680 1710 30 2 2.1 3060 3070 10 0 7.2 5770 5780 10 0 14.2 7090 7020 -70 -1 Control 28.1 7840 7890 50 1 30 1 0.9 330 100 -230 -70 1.9 1340 950 -390 -29 3.4 2450 1870 -580 -24 12.3 4960 4470 -490 -10 24.9 6740 6050 -690 -10 Cold 49.0 8560 7660 -900 -11 550 26 0.3 610 1570 960 157 0.8 2330 3950 1620 70 1.4 3240 5250 2010 62 5.0 6060 7550 1490 25 10.1 6490 8260 1770 27 Hot 19.8 6540 8660 2120 32 1660 62 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-17: Estimated strength versus actual strength for 50% Slag mixture 246 Table D-18: Error using AM function based on ASTM C 1074 methods for the Type III - 0.44 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 730 790 60 8 1.0 3040 2850 -190 -6 2.0 3820 3890 70 2 7.2 5100 5350 250 5 14.0 5780 5790 10 0 Control 28.2 6240 6050 -190 -3 130 4 1.0 750 970 220 29 1.8 2370 2230 -140 -6 3.4 3660 3270 -390 -11 12.1 5630 5050 -580 -10 24.9 6430 5650 -780 -12 Cold 49.2 7070 5980 -1090 -15 530 14 0.3 2400 2070 -330 -14 0.7 3210 3690 480 15 1.4 3740 4590 850 23 5.3 5100 5750 650 13 10.0 5290 6010 720 14 Hot 20.2 5500 6130 630 11 610 15 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-18: Estimated strength versus actual strength for Type III - 0.44 mixture 247 Table D-19: Error using AM function based on ASTM C 1074 methods for the 70/20/10 - 0.37 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1300 1530 230 18 1.0 3650 3340 -310 -8 2.1 5240 5050 -190 -4 7.0 7670 8310 640 8 14.1 10110 9780 -330 -3 Control 28.1 10800 10760 -40 0 290 7 0.9 1040 1650 610 59 1.8 3060 3100 40 1 3.5 4650 4480 -170 -4 12.2 7000 7870 870 12 25.0 8870 9540 670 8 Cold 49.2 10820 10600 -220 -2 430 14 0.3 1640 2290 650 40 0.7 3540 4980 1440 41 1.4 5020 6800 1780 35 4.9 7390 9660 2270 31 9.8 8240 10710 2470 30 Hot 20.0 8490 11180 2690 32 1880 35 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-19: Estimated strength versus actual strength for 70/20/10 - 0.37 mixture 248 Table D-20: Error using AM function based on ASTM C 1074 methods for the 70/20/10 - 0.44 mixture Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 510 630 120 24 1.0 1770 1620 -150 -8 2.0 2610 2540 -70 -3 7.2 4680 4900 220 5 13.9 6250 6120 -130 -2 Control 27.9 7140 7150 10 0 120 7 1.0 370 730 360 97 1.8 1450 1260 -190 -13 3.4 2560 2020 -540 -21 12.1 4510 4380 -130 -3 24.9 5820 5810 -10 0 Cold 48.9 7360 6900 -460 -6 280 23 0.3 980 1400 420 43 0.8 1860 2780 920 49 1.5 2720 4030 1310 48 5.0 4500 6290 1790 40 10.2 4840 7290 2450 51 Hot 19.9 5300 7860 2560 48 1580 47 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure D-20: Estimated strength versus actual strength for 70/20/10 - 0.44 mixture 249 APPENDIX E: STRENGTH-MATURITY RELATIONSHIPS FOR NURSE-SAUL MATURITY FUNCTION, MODIFIED ASTM METHOD 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 10,000 12,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-1: Strength-maturity plot, Modified ASTM method for Type I - 0.44 To = 17?F 250 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-2: Strength-maturity plot, Modified ASTM method for Type I - 0.48 mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-3: Strength-maturity plot, Modified ASTM method for 30% F mixture To = 28?F To = 39?F 251 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Realtionship Control Data Cold Data Hot Data Figure E-4: Strength-maturity plot, Modified ASTM method for 20% C mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-5: Strength-maturity plot, Modified ASTM method for 30% C mixture To = 46?F To = 35?F 252 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-6: Strength-maturity plot, Modified ASTM method for 30% Slag mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-7: Strength-maturity plot, Modified ASTM method for 50% Slag mixture To = 40?F To = 25?F 253 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 2,000 4,000 6,000 8,000 10,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-8: Strength-maturity plot, Modified ASTM method for Type III - 0.44 mixture 0 2,000 4,000 6,000 8,000 10,000 12,000 0 5,000 10,000 15,000 20,000 25,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-9: Strength-maturity plot, Modified ASTM method for 70/20/10 - 0.37 mixture To = 29?F To = -15?F 254 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 5,000 10,000 15,000 20,000 25,000 Maturity Index (?F?hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure E-10: Strength-maturity plot, Modified ASTM method for 70/20/10-0.44 mixture To = -17?F 255 APPENDIX F: STRENGTH-MATURITY RELATIONSHIPS FOR ARRHENIUS MATURITY FUNCTION, MODIFIED ASTM METHOD 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-1: Strength-maturity plot, Modified ASTM method for Type I - 0.44 mixture E = 25,000 J/mol 256 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-2: Strength-maturity plot, Modified ASTM method for Type I - 0.48 mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 300 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-3: Strength-maturity plot, Modified ASTM method for 30% F mixture E = 30,800 J/mol E = 40,200 J/mol 257 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 300 350 400 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-4: Strength-maturity plot, Modified ASTM method for 20% C mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 300 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-5: Strength-maturity plot, Modified ASTM method for 30% C mixture E = 54,100 J/mol E = 35,300 J/mol 258 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 300 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-6: Strength-maturity plot, Modified ASTM method for 30% Slag mixture 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 0 50 100 150 200 250 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-7: Strength-maturity plot, Modified ASTM method for 50% Slag mixture E = 39,000 J/mol E = 28,600 J/mol 259 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 300 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-8: Strength-maturity plot, Modified ASTM method for Type III - 0.44 mixture 0 2,000 4,000 6,000 8,000 10,000 12,000 0 50 100 150 200 250 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-9: Strength-maturity plot, Modified ASTM method for 70/20/10 - 0.37 mixture E = 38,100 J/mol E = 16,700 J/mol 260 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 50 100 150 200 250 Equivalent Age Maturity (hr) Compressive Strength (psi) S-M Relationship Control Data Cold Data Hot Data Figure F-10: Strength-maturity plot, Modified ASTM method for 70/20/10-0.44 mixture E = 14,800 J/mol 261 APPENDIX G: ERROR TABLES FOR MODIFIED ASTM METHOD Table G-1: Error using NSM function based on Modified ASTM method for Type I - 0.44 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 830 990 160 19 850 20 2 1.0 1,990 1,800 -190 -10 1,920 -70 -4 2.0 2,840 2,710 -130 -5 2,900 60 2 Control 7.3 4,210 4,470 260 6 190 10 4,190 -20 0 40 2 0.9 420 800 380 90 940 520 124 1.9 2,050 1,660 -390 -19 2,210 160 8 3.5 3,090 2,410 -680 -22 3,070 -20 -1 Cold 12.4 4,580 3,980 -600 -13 510 36 4,210 -370 -8 270 35 0.3 1,270 930 -340 -27 540 -730 -57 0.9 2,570 2,300 -270 -11 2,270 -300 -12 1.5 3,390 3,160 -230 -7 3,090 -300 -9 Hot 5.0 4,380 4,680 300 7 290 13 4,210 -170 -4 380 20 262 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-1: Error plot for the Type I - 0.44 mixture, Modified method, NSM 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-2: Error plot for the Type I - 0.44 mixture, ASTM C 1074 method, NSM 263 Table G-2: Error using NSM function based on Modified ASTM method for Type I - 0.48 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 630 780 150 24 660 30 5 1.0 1,710 1,540 -170 -10 1,630 -80 -5 2.0 2,570 2,440 -130 -5 2,630 60 2 Control 7.4 4,120 4,260 140 3 150 11 4,110 -10 0 50 3 0.9 410 580 170 41 670 260 63 1.9 1,600 1,210 -390 -24 1,650 50 3 3.6 3,040 1,770 -1,270 -42 2,450 -590 -19 Cold 12.0 4,610 3,170 -1,440 -31 820 35 3,840 -770 -17 420 26 0.3 850 780 -70 -8 550 -300 -35 0.8 2,180 1,980 -200 -9 1,990 -190 -9 1.4 2,770 2,740 -30 -1 2,760 -10 0 Hot 5.1 4,010 4,530 520 13 210 8 4,190 180 4 170 12 264 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-3: Error plot for the Type I - 0.48 mixture, Modified method, NSM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-4: Error plot for the Type I - 0.48 mixture, ASTM C 1074 method, NSM 265 Table G-3: Error using NSM function based on Modified ASTM method for 30% F mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,180 1,220 40 3 1,230 50 4 1.1 2,200 1,910 -290 -13 2,080 -120 -5 2.0 2,860 2,810 -50 -2 2,960 100 3 Control 7.0 4,490 4,810 320 7 180 6 4,470 -20 0 70 3 0.9 670 1,250 580 87 840 170 25 1.9 1,990 2,170 180 9 1,800 -190 -10 3.5 3,090 3,010 -80 -3 2,470 -620 -20 Cold 12.2 4,990 4,860 -130 -3 240 25 3,870 -1,120 -22 530 19 0.4 1,350 1,220 -130 -10 1,500 150 11 0.8 2,860 2,180 -680 -24 2,670 -190 -7 1.5 3,620 3,080 -540 -15 3,500 -120 -3 Hot 5.0 4,960 5,000 40 1 350 12 4,770 -190 -4 160 6 266 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-5: Error plot for the 30% F mixture, Modified method, NSM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-6: Error plot for the 30% F mixture, ASTM C 1074 method, NSM 267 Table G-4: Error using NSM function based on Modified ASTM method for 20% C mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,220 1,200 -20 -2 1,160 -60 -5 1.0 2,210 2,320 110 5 2,390 180 8 2.0 3,730 3,540 -190 -5 3,590 -140 -4 Control 7.0 5,330 5,440 110 2 110 3 5,360 30 1 100 4 1.2 1,040 1,010 -30 -3 110 -930 -89 1.8 1,760 1,810 50 3 900 -860 -49 3.4 3,140 2,610 -530 -17 1,260 -1,880 -60 Cold 12.0 5,410 4,590 -820 -15 360 9 2,930 -2,480 -46 1540 61 0.5 1,930 1,780 -150 -8 2,060 130 7 0.8 3,600 3,040 -560 -16 3,380 -220 -6 1.4 3,900 4,030 130 3 4,310 410 11 Hot 5.0 5,360 5,740 380 7 310 8 5,760 400 7 290 8 268 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-7: Error plot for the 20% C mixture, Modified method, NSM 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-8: Error plot for the 20% C mixture, ASTM C 1074 method, NSM 269 Table G-5: Error using NSM function based on Modified ASTM method for 30% C mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 670 790 120 18 720 50 7 1.0 1,940 1,790 -150 -8 1,830 -110 -6 2.0 3,060 3,020 -40 -1 3,140 80 3 Control 7.2 5,250 5,260 10 0 80 7 5,230 -20 0 70 4 1.3 570 330 -240 -42 640 70 12 1.9 1,680 920 -760 -45 1,450 -230 -14 3.5 2,620 1,750 -870 -33 2,560 -60 -2 Cold 12.1 4,650 4,460 -190 -4 520 31 4,980 330 7 170 9 0.4 1,050 1,140 90 9 970 -80 -8 0.8 2,710 2,550 -160 -6 2,460 -250 -9 1.4 3,900 3,700 -200 -5 3,610 -290 -7 Hot 5.0 5,870 5,880 10 0 120 5 5,620 -250 -4 220 7 270 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% error 20% Error Figure G-9: Error plot for the 30% C mixture, Modified method, NSM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot 10% Error 20% Error Figure G-10: Error plot for the 30% C mixture, ASTM C 1074 method, NSM 271 Table G-6: Error using NSM function based on Modified ASTM method for 30% Slag mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,430 1,500 70 5 1,470 40 3 1.0 2,730 2,580 -150 -5 2,600 -130 -5 2.0 3,650 3,700 50 1 3,750 100 3 Control 7.1 5,770 5,780 10 0 70 3 5,760 -10 0 70 3 0.9 610 640 30 5 890 280 46 1.9 2,050 1,450 -600 -29 1,980 -70 -3 3.6 3,400 2,040 -1,360 -40 2,840 -560 -16 Cold 12.2 5,080 3,590 -1,490 -29 870 26 4,720 -360 -7 320 18 0.4 1,500 1,760 260 17 1,620 120 8 0.8 3,130 3,260 130 4 3,140 10 0 2.9 4,990 5,640 650 13 5,480 490 10 Hot 4.9 5,240 6,300 1,060 20 530 14 6,110 870 17 370 9 272 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-11: Error plot for the 30% Slag mixture, Modified method, NSM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-12: Error plot for the 30% Slag mixture, ASTM C 1074 method, NSM 273 Table G-7: Error using NSM function based on Modified ASTM method for 50% Slag mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 760 740 -20 -3 760 0 0 1.0 1,680 1,690 10 1 1,670 -10 -1 2.1 3,060 3,070 10 0 3,070 10 0 Control 7.2 5,770 5,790 20 0 20 1 5,770 0 0 10 0 0.9 330 0 -330 -100 600 270 82 1.9 1,340 280 -1,060 -79 1,710 370 28 3.4 2,450 740 -1,710 -70 2,840 390 16 Cold 12.3 4,960 2,010 -2,950 -59 1510 77 5,480 520 10 390 34 0.3 610 770 160 26 500 -110 -18 0.8 2,330 2,660 330 14 2,100 -230 -10 1.4 3,240 3,900 660 20 3,230 -10 0 Hot 5.0 6,060 6,710 650 11 450 18 6,050 -10 0 90 7 274 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-13: Error plot for the 50% Slag mixture, Modified method, NSM 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-14: Error plot for the 50% Slag mixture, ASTM C 1074 method, NSM 275 Table G-8: Error using NSM function based on Modified ASTM method for Type III - 0.44 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 730 800 70 10 750 20 3 1.0 3,040 2,820 -220 -7 2,950 -90 -3 2.0 3,820 3,920 100 3 3,950 130 3 Control 7.2 5,100 5,370 270 5 170 6 5,060 -40 -1 70 2 1.0 750 980 230 31 1,360 610 81 1.8 2,370 2,230 -140 -6 2,730 360 15 3.4 3,660 3,090 -570 -16 3,610 -50 -1 Cold 12.1 5,630 4,800 -830 -15 440 17 4,890 -740 -13 440 28 0.3 2,400 1,300 -1,100 -46 1,210 -1,190 -50 0.7 3,210 3,020 -190 -6 3,040 -170 -5 1.4 3,740 4,110 370 10 4,010 270 7 Hot 5.3 5,100 5,560 460 9 530 18 5,150 50 1 420 16 276 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-15: Error plot for the Type III - 0.44 mixture, Modified method, NSM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-16: Error plot for the Type III - 0.44 mixture, ASTM C 1074 method, NSM 277 Table G-9: Error using NSM function based on Modified ASTM method for 70/20/10 - 0.37 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,300 1,550 250 19 1,340 40 3 1.0 3,650 3,320 -330 -9 3,530 -120 -3 2.1 5,240 5,050 -190 -4 5,350 110 2 Control 7.0 7,670 8,320 650 8 360 10 7,640 -30 0 80 2 0.9 1,040 1,520 480 46 2,330 1,290 124 1.8 3,060 2,810 -250 -8 4,400 1,340 44 3.5 4,650 3,790 -860 -18 5,890 1,240 27 Cold 12.2 7,000 6,700 -300 -4 470 19 7,970 970 14 1210 52 0.3 1,640 1,650 10 1 740 -900 -55 0.7 3,540 4,010 470 13 3,400 -140 -4 1.4 5,020 5,820 800 16 5,080 60 1 Hot 4.9 7,390 9,090 1,700 23 750 13 7,560 170 2 320 16 278 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-17: Error plot for the 70/20/10 - 0.37 mixture, Modified method, NSM 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-18: Error plot for the 70/20/10 - 0.37 mixture, ASTM C 1074 method, NSM 279 Table G-10: Error using NSM function based on Modified ASTM method for the 70/20/10 - 0.44 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 510 650 140 27 560 50 10 1.0 1,770 1,600 -170 -10 1,650 -120 -7 2.0 2,610 2,540 -70 -3 2,700 90 3 Control 7.2 4,680 4,910 230 5 150 11 4,660 -20 0 70 5 1.0 370 620 250 68 1,300 930 251 1.8 1,450 1,070 -380 -26 2,100 650 45 3.4 2,560 1,500 -1,060 -41 3,140 580 23 Cold 12.1 4,510 3,150 -1,360 -30 760 41 4,970 460 10 660 82 0.3 980 920 -60 -6 520 -460 -47 0.8 1,860 2,050 190 10 1,600 -260 -14 1.5 2,720 3,230 510 19 2,640 -80 -3 Hot 5.0 4,500 5,700 1,200 27 490 15 4,570 70 2 220 16 280 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-19: Error plot for the 70/20/10 - 0.44 mixture, Modified method, NSM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-20: Error plot for the 70/20/10 - 0.44 mixture, ASTM C 1074 method, NSM 281 Table G-11: Error using AM function based on Modified ASTM method for Type I - 0.44 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 830 980 150 18 850 20 2 1.0 1,990 1,800 -190 -10 1,920 -70 -4 2.0 2,840 2,710 -130 -5 2,900 60 2 Contro l 7.3 4,210 4,460 250 6 180 10 4,190 -20 0 40 2 0.9 420 900 480 114 950 530 126 1.9 2,050 1,830 -220 -11 2,230 180 9 3.5 3,090 2,690 -400 -13 3,110 20 1 Cold 12.4 4,580 4,380 -200 -4 330 36 4,260 -320 -7 260 36 0.3 1,270 1,150 -120 -9 740 -530 -42 0.9 2,570 2,610 40 2 2,470 -100 -4 1.5 3,390 3,450 60 2 3,250 -140 -4 Hot 5.0 4,380 4,830 450 10 170 6 4,280 -100 -2 220 13 282 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-21: Error plot for the Type I - 0.44 mixture, Modified method, AM 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-22: Error plot for the Type I - 0.44 mixture, ASTM C 1074 method, AM 283 Table 6.12: Error using AM function based on Modified ASTM method for Type I - 0.48 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 630 780 150 24 650 20 3 1.0 1,710 1,550 -160 -9 1,640 -70 -4 2.0 2,570 2,440 -130 -5 2,630 60 2 Control 7.4 4,120 4,250 130 3 140 10 4,110 -10 0 40 2 0.9 410 670 260 63 740 330 80 1.9 1,600 1,380 -220 -14 1,760 160 10 3.6 3,040 2,110 -930 -31 2,610 -430 -14 Cold 12.0 4,610 3,810 -800 -17 550 31 4,010 -600 -13 380 29 0.3 850 1,000 150 18 710 -140 -16 0.8 2,180 2,450 270 12 2,250 70 3 1.4 2,770 3,180 410 15 2,970 200 7 Hot 5.1 4,010 4,810 800 20 410 16 4,300 290 7 180 9 284 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-23: Error plot for the Type I - 0.48 mixture, Modified method, AM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-24: Error plot for the Type I - 0.48 mixture, ASTM C 1074 method, AM 28 5 Table G-13: Error using AM function based on Modified ASTM method for 30% F mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,180 1,220 40 3 1,230 50 4 1.1 2,200 1,910 -290 -13 2,080 -120 -5 2.0 2,860 2,810 -50 -2 2,950 90 3 Control 7.0 4,490 4,810 320 7 180 6 4,470 -20 0 70 3 0.9 670 1,320 650 97 1,040 370 55 1.9 1,990 2,280 290 15 2,080 90 5 3.5 3,090 3,170 80 3 2,890 -200 -6 Cold 12.2 4,990 5,040 50 1 270 29 4,370 -620 -12 320 20 0.4 1,350 1,290 -60 -4 1,870 520 39 0.8 2,860 2,280 -580 -20 3,050 190 7 1.5 3,620 3,190 -430 -12 3,820 200 6 Hot 5.0 4,960 5,080 120 2 300 10 4,920 -40 -1 238 13 286 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-25: Error plot for the 30% F mixture, Modified method, AM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-26: Error plot for the 30% F mixture, ASTM C 1074 method, AM 287 Table G-14: Error using AM function based on Modified ASTM method for 20% C mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,220 1,200 -20 -2 1,150 -70 -6 1.0 2,210 2,330 120 5 2,400 190 9 2.0 3,730 3,540 -190 -5 3,570 -160 -4 Control 7.0 5,330 5,440 110 2 110 4 5,360 30 1 110 5 1.2 1,040 1,310 270 26 800 -240 -23 1.8 1,760 2,140 380 22 1,630 -130 -7 3.4 3,140 3,230 90 3 2,660 -480 -15 Cold 12.0 5,410 5,240 -170 -3 230 13 4,770 -640 -12 370 14 0.5 1,930 2,210 280 15 2,970 1,040 54 0.8 3,600 3,530 -70 -2 4,280 680 19 1.4 3,900 4,420 520 13 4,980 1,080 28 Hot 5.0 5,360 5,910 550 10 360 10 6,050 690 13 870 28 288 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-27: Error plot for the 20% C mixture, Modified method, AM 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-28: Error plot for the 20% C mixture, ASTM C 1074 method, AM 289 Table G-15: Error using AM function based on Modified ASTM method for 30% C mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 670 780 110 16 720 50 7 1.0 1,940 1,800 -140 -7 1,830 -110 -6 2.0 3,060 3,020 -40 -1 3,140 80 3 Contr ol 7.2 5,250 5,260 10 0 80 6 5,230 -20 0 70 4 1.3 570 740 170 30 980 410 72 1.9 1,680 1,480 -200 -12 1,860 180 11 3.5 2,620 2,620 0 0 3,090 470 18 Cold 12.1 4,650 5,160 510 11 220 13 5,310 660 14 430 29 0.4 1,050 1,610 560 53 1,270 220 21 0.8 2,710 3,250 540 20 2,880 170 6 1.4 3,900 4,300 400 10 3,960 60 2 Hot 5.0 5,870 6,200 330 6 460 22 5,780 -90 -2 140 8 290 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-29: Error plot for the 30% C mixture, Modified method, AM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-30: Error plot for the 30% C mixture, ASTM C 1074 method, AM 291 Table G-16: Error using AM function based on Modified ASTM method for 30% Slag mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,430 1,490 60 4 1,470 40 3 1.0 2,730 2,610 -120 -4 2,600 -130 -5 2.0 3,650 3,700 50 1 3,750 100 3 Control 7.1 5,770 5,770 0 0 60 2 5,760 -10 0 70 3 0.9 610 840 230 38 1,180 570 93 1.9 2,050 1,900 -150 -7 2,440 390 19 3.6 3,400 2,910 -490 -14 3,550 150 4 Cold 12.2 5,080 5,100 20 0 220 15 5,590 510 10 410 32 0.4 1,500 2,660 1,160 77 2,010 510 34 0.8 3,130 4,250 1,120 36 3,560 430 14 2.9 4,990 6,260 1,270 25 5,740 750 15 Hot 4.9 5,240 6,730 1,490 28 1260 42 6,290 1,050 20 690 21 292 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-31: Error plot for the 30% Slag mixture, Modified method, AM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-32: Error plot for the 30% Slag mixture, ASTM C 1074 method, AM 293 Table G-17: Error using AM function based on Modified ASTM method for 50% Slag mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 760 730 -30 -4 760 0 0 1.0 1,680 1,710 30 2 1,670 -10 -1 2.1 3,060 3,070 10 0 3,070 10 0 Control 7.2 5,770 5,780 10 0 20 2 5,770 0 0 10 0 0.9 330 100 -230 -70 670 340 103 1.9 1,340 950 -390 -29 1,850 510 38 3.4 2,450 1,870 -580 -24 3,070 620 25 Cold 12.3 4,960 4,470 -490 -10 420 33 5,810 850 17 580 46 0.3 610 1,570 960 157 670 60 10 0.8 2,330 3,950 1,620 70 2,380 50 2 1.4 3,240 5,250 2,010 62 3,540 300 9 Hot 5.0 6,060 7,550 1,490 25 1520 78 6,280 220 4 160 6 294 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-33: Error plot for the 50% Slag mixture, Modified method, AM 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-34: Error plot for the 50% Slag mixture, ASTM C 1074 method, AM 295 Table G-18: Error using AM function based on Modified ASTM method for Type III - 0.44 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 730 790 60 8 740 10 1 1.0 3,040 2,850 -190 -6 2,960 -80 -3 2.0 3,820 3,890 70 2 3,930 110 3 Control 7.2 5,100 5,350 250 5 140 5 5,070 -30 -1 60 2 1.0 750 970 220 29 1,170 420 56 1.8 2,370 2,230 -140 -6 2,560 190 8 3.4 3,660 3,270 -390 -11 3,550 -110 -3 Cold 12.1 5,630 5,050 -580 -10 330 14 4,920 -710 -13 360 20 0.3 2,400 2,070 -330 -14 1,980 -420 -18 0.7 3,210 3,690 480 15 3,590 380 12 1.4 3,740 4,590 850 23 4,380 640 17 Hot 5.3 5,100 5,750 650 13 580 16 5,300 200 4 410 13 296 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-35: Error plot for the Type III - 0.44 mixture, Modified method, AM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-36: Error plot for the Type III - 0.44 mixture, ASTM C 1074 method, AM 297 Table G-19: Error using AM function based on Modified ASTM method for 70/20/10 - 0.37 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,300 1,530 230 18 1,340 40 3 1.0 3,650 3,340 -310 -8 3,530 -120 -3 2.1 5,240 5,050 -190 -4 5,350 110 2 Control 7.0 7,670 8,310 640 8 340 10 7,640 -30 0 80 2 0.9 1,040 1,650 610 59 2,300 1,260 121 1.8 3,060 3,100 40 1 4,370 1,310 43 3.5 4,650 4,480 -170 -4 5,870 1,220 26 Cold 12.2 7,000 7,870 870 12 420 19 7,970 970 14 1190 51 0.3 1,640 2,290 650 40 910 -730 -45 0.7 3,540 4,980 1,440 41 3,610 70 2 1.4 5,020 6,800 1,780 35 5,270 250 5 Hot 4.9 7,390 9,660 2,270 31 1540 37 7,660 270 4 330 14 298 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-37: Error plot for the 70/20/10 - 0.37 mixture, Modified method, AM 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-38: Error plot for the 70/20/10 - 0.37 mixture, ASTM C 1074 method, AM 299 Table G-20: Error using AM function based on Modified ASTM method for 70/20/10 - 0.44 mixture ASTM C 1074 Modified ASTM Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 510 630 120 24 560 50 10 1.0 1,770 1,620 -150 -8 1,650 -120 -7 2.0 2,610 2,540 -70 -3 2,700 90 3 Control 7.2 4,680 4,900 220 5 140 10 4,660 -20 0 70 5 1.0 370 730 360 97 1,310 940 254 1.8 1,450 1,260 -190 -13 2,110 660 46 3.4 2,560 2,020 -540 -21 3,170 610 24 Cold 12.1 4,510 4,380 -130 -3 310 34 5,000 490 11 680 84 0.3 980 1,400 420 43 560 -420 -43 0.8 1,860 2,780 920 49 1,660 -200 -11 1.5 2,720 4,030 1,310 48 2,710 -10 0 Hot 5.0 4,500 6,290 1,790 40 1110 45 4,620 120 3 190 14 300 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-39: Error plot for the 70/20/10 - 0.44 mixture, Modified method, AM 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure G-40: Error plot for the 70/20/10 - 0.44 mixture, ASTM C 1074 method, AM 301 APPENDIX H: ERROR TABLES FOR NURSE-SAUL MATURITY FUNCTION, SIMPLIFIED METHOD Table H-1: Error using NSM function based on simplified method for Type I - 0.44 mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 830 850 20 2 840 10 1 1.0 1,990 1,920 -70 -4 1,930 -60 -3 2.0 2,840 2,900 60 2 2,890 50 2 Control 7.3 4,210 4,190 -20 0 40 2 4,200 -10 0 30 2 0.9 420 970 550 131 650 230 55 1.9 2,050 2,240 190 9 1,890 -160 -8 3.5 3,090 3,100 10 0 2,740 -350 -11 Cold 12.4 4,580 4,230 -350 -8 280 37 3,990 -590 -13 330 22 0.3 1,270 520 -750 -59 700 -570 -45 0.9 2,570 2,250 -320 -12 2,440 -130 -5 1.5 3,390 3,070 -320 -9 3,230 -160 -5 Hot 5.0 4,380 4,200 -180 -4 390 21 4,300 -80 -2 240 14 302 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-1: Error plot for the Type I - 0.44 mixture, simplified method, To = 14?F 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-2: Error plot for the Type I - 0.44 mixture, simplified method, To = 32?F 303 Table H-2: Error using NSM function based on simplified method for Type I - 0.48 mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 630 660 30 5 650 20 3 1.0 1,710 1,630 -80 -5 1,640 -70 -4 2.0 2,570 2,640 70 3 2,630 60 2 Control 7.4 4,120 4,100 -20 0 50 3 4,110 -10 0 40 2 0.9 410 880 470 115 570 160 39 1.9 1,600 1,920 320 20 1,530 -70 -4 3.6 3,040 2,750 -290 -10 2,300 -740 -24 Cold 12.0 4,610 4,070 -540 -12 410 39 3,710 -900 -20 468 22 0.3 850 470 -380 -45 580 -270 -32 0.8 2,180 1,860 -320 -15 2,040 -140 -6 1.4 2,770 2,640 -130 -5 2,810 40 1 Hot 5.1 4,010 4,120 110 3 240 17 4,230 220 5 168 11 304 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-3: Error plot for the Type I - 0.48 mixture, simplified method, To = 14?F 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-4: Error plot for the Type I - 0.48 mixture, simplified method, To = 32?F 305 Table H-3: Error using NSM function based on simplified method for 30% F mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,180 1,240 60 5 1,230 50 4 1.1 2,200 2,060 -140 -6 2,070 -130 -6 2.0 2,860 2,970 110 4 2,960 100 3 Control 7.0 4,490 4,470 -20 0 80 4 4,470 -20 0 80 4 0.9 670 1,320 650 97 1,040 370 55 1.9 1,990 2,410 420 21 2,070 80 4 3.5 3,090 3,210 120 4 2,820 -270 -9 Cold 12.2 4,990 4,550 -440 -9 410 33 4,250 -740 -15 370 21 0.4 1,350 1,190 -160 -12 1,380 30 2 0.8 2,860 2,310 -550 -19 2,530 -330 -12 1.5 3,620 3,170 -450 -12 3,370 -250 -7 Hot 5.0 4,960 4,570 -390 -8 390 13 4,690 -270 -5 220 7 306 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-5: Error plot for the 30% F mixture, simplified method, To = 14?F 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-6: Error plot for the 30% F mixture, simplified method, To = 32?F 307 Table H-4: Error using NSM function based on simplified method for 20% C mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,220 1,170 -50 -4 1,160 -60 -5 1.0 2,210 2,350 140 6 2,360 150 7 2.0 3,730 3,620 -110 -3 3,610 -120 -3 Control 7.0 5,330 5,350 20 0 80 3 5,360 30 1 90 4 1.2 1,040 1,670 630 61 1,190 150 14 1.8 1,760 2,590 830 47 2,100 340 19 3.4 3,140 3,700 560 18 3,070 -70 -2 Cold 12.0 5,410 5,370 -40 -1 520 32 4,940 -470 -9 260 11 0.5 1,930 1,440 -490 -25 1,690 -240 -12 0.8 3,600 2,730 -870 -24 3,000 -600 -17 1.4 3,900 3,750 -150 -4 3,980 80 2 Hot 5.0 5,360 5,420 60 1 390 14 5,560 200 4 280 9 308 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-7: Error plot for the 20% C mixture, simplified method, To = 14?F 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-8: Error plot for the 20% C mixture, simplified method, To = 32?F 309 Table H-5: Error using NSM function based on simplified method for 30% C mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 670 730 60 9 720 50 7 1.0 1,940 1,800 -140 -7 1,830 -110 -6 2.0 3,060 3,160 100 3 3,150 90 3 Control 7.2 5,250 5,230 -20 0 80 5 5,230 -20 0 70 4 1.3 570 1,310 740 130 790 220 39 1.9 1,680 2,200 520 31 1,620 -60 -4 3.5 2,620 3,400 780 30 2,770 150 6 Cold 12.1 4,650 5,410 760 16 700 52 5,100 450 10 220 14 0.4 1,050 690 -360 -34 920 -130 -12 0.8 2,710 2,090 -620 -23 2,390 -320 -12 1.4 3,900 3,250 -650 -17 3,540 -360 -9 Hot 5.0 5,870 5,360 -510 -9 540 21 5,570 -300 -5 280 10 310 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-9: Error plot for the 30% C mixture, simplified method, To = 14?F 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-10: Error plot for the 30% C mixture, simplified method, To = 32?F 311 Table H-6: Error using NSM function based on simplified method for 30% Slag mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,430 1,490 60 4 1,480 50 3 1.0 2,730 2,570 -160 -6 2,590 -140 -5 2.0 3,650 3,770 120 3 3,760 110 3 Control 7.1 5,770 5,750 -20 0 90 3 5,750 -20 0 80 3 0.9 610 1,560 950 156 1,210 600 98 1.9 2,050 2,900 850 41 2,430 380 19 3.6 3,400 4,010 610 18 3,460 60 2 Cold 12.2 5,080 5,840 760 15 790 58 5,400 320 6 340 31 0.4 1,500 1,320 -180 -12 1,490 -10 -1 0.8 3,130 2,730 -400 -13 2,960 -170 -5 2.9 4,990 5,110 120 2 5,320 330 7 Hot 4.9 5,240 5,800 560 11 320 9 5,980 740 14 310 7 312 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-11: Error plot for the 30% Slag mixture, simplified method, To = 14?F 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-12: Error plot for the 30% Slag mixture, simplified method, To = 32?F 313 Table H-7: Error using NSM function based on simplified method for 50% Slag mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 760 760 0 0 760 0 0 1.0 1,680 1,660 -20 -1 1,670 -10 -1 2.1 3,060 3,070 10 0 3,060 0 0 Control 7.2 5,770 5,770 0 0 10 0 5,770 0 0 0 0 0.9 330 760 430 130 450 120 36 1.9 1,340 1,960 620 46 1,470 130 10 3.4 2,450 3,170 720 29 2,500 50 2 Cold 12.3 4,960 5,830 870 18 660 56 5,050 90 2 100 12 0.3 610 440 -170 -28 560 -50 -8 0.8 2,330 1,970 -360 -15 2,210 -120 -5 1.4 3,240 3,060 -180 -6 3,370 130 4 Hot 5.0 6,060 5,890 -170 -3 220 13 6,190 130 2 110 5 314 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-13: Error plot for the 50% Sla g mixture, simplified method, To = 14?F 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-14: Error plot for the 50% Slag mixture, simplified method, To = 32?F 315 Table H-8: Error using NSM function based on simplified method for Type III - 0.44 mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 730 750 20 3 750 20 3 1.0 3,040 2,940 -100 -3 2,950 -90 -3 2.0 3,820 3,960 140 4 3,940 120 3 Control 7.2 5,100 5,050 -50 -1 80 3 5,060 -40 -1 70 2 1.0 750 1,780 1,030 137 1,240 490 65 1.8 2,370 3,100 730 31 2,620 250 11 3.4 3,660 3,970 310 8 3,490 -170 -5 Cold 12.1 5,630 5,050 -580 -10 660 47 4,820 -810 -14 430 24 0.3 2,400 1,000 -1,400 -58 1,250 -1,150 -48 0.7 3,210 2,890 -320 -10 3,080 -130 -4 1.4 3,740 3,900 160 4 4,040 300 8 Hot 5.3 5,100 5,090 -10 0 470 18 5,170 70 1 410 15 316 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-15: Error plot for the Type III - 0.44 mixture, simplified method, To = 14?F 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-16: Error plot for the Type III - 0.44 mixture, simplified method, To = 32?F 317 Table H-9: Error using NSM function based on simplified method for 70/20/10 - 0.37 mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,300 1,330 30 2 1,330 30 2 1.0 3,650 3,550 -100 -3 3,570 -80 -2 2.1 5,240 5,340 100 2 5,310 70 1 Control 7.0 7,670 7,640 -30 0 70 2 7,650 -20 0 50 2 0.9 1,040 2,030 990 95 1,600 560 54 1.8 3,060 4,040 980 32 3,460 400 13 3.5 4,650 5,480 830 18 4,760 110 2 Cold 12.2 7,000 7,750 750 11 890 39 7,240 240 3 330 18 0.3 1,640 990 -650 -40 1,290 -350 -21 0.7 3,540 3,720 180 5 4,090 550 16 1.4 5,020 5,380 360 7 5,730 710 14 Hot 4.9 7,390 7,740 350 5 390 14 7,960 570 8 550 15 318 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-17: Error plot for the 70/20/10 - 0.37 mixture, simplified method, To = 14?F 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-18: Error plot for the 70/20/10 - 0.37 mixture, simplified method, To = 32?F 319 Table H-10: Error using NSM function based on simplified method for the 70/20/10 - 0.44 mixture To = 14?F (-10?C) To = 32?F (0?C) Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 510 550 40 8 550 40 8 1.0 1,770 1,660 -110 -6 1,680 -90 -5 2.0 2,610 2,690 80 3 2,680 70 3 Control 7.2 4,680 4,660 -20 0 60 4 4,670 -10 0 50 4 1.0 370 1,090 720 195 820 450 122 1.8 1,450 1,850 400 28 1,510 60 4 3.4 2,560 2,820 260 10 2,320 -240 -9 Cold 12.1 4,510 4,760 250 6 410 59 4,300 -210 -5 240 35 0.3 980 630 -350 -36 750 -230 -23 0.8 1,860 1,780 -80 -4 1,970 110 6 1.5 2,720 2,850 130 5 3,080 360 13 Hot 5.0 4,500 4,760 260 6 210 13 4,970 470 10 290 13 320 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-19: Error plot for the 70/20/10 - 0.44 mixture, simplified method, To = 14?F 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure H-20: Error plot for the 70/20/10 - 0.44 mixture, simplified method, To = 32?F 321 APPENDIX I: ERROR TABLES FOR ARRHENIUS MATURITY FUNCTION, SIMPLIFIED METHOD Table I-1: Error using NSM function based on simplified method for Type I - 0.44 mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 830 840 10 1 850 20 2 1.0 1,990 1,940 -50 -3 1,920 -70 -4 2.0 2,840 2,880 40 1 2,900 60 2 Control 7.3 4,210 4,200 -10 0 30 1 4,190 -20 0 40 2 0.9 420 580 160 38 950 530 126 1.9 2,050 1,820 -230 -11 2,230 180 9 3.5 3,090 2,730 -360 -12 3,110 20 1 Cold 12.4 4,580 4,080 -500 -11 310 18 4,260 -320 -7 260 36 0.3 1,270 1,270 0 0 750 -520 -41 0.9 2,570 2,960 390 15 2,470 -100 -4 1.5 3,390 3,630 240 7 3,250 -140 -4 Hot 5.0 4,380 4,480 100 2 180 6 4,280 -100 -2 220 13 322 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-1: Error plot for the Type I - 0.44 mixture, simplified method, E= 40 kJ/mol 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-2: Error plot for the Type I - 0.44 mixture, simplified method, E = 25 kJ/mol 323 Table I-2: Error using NSM function based on simplified method for Type I - 0.48 mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 630 650 20 3 660 30 5 1.0 1,710 1,650 -60 -4 1,630 -80 -5 2.0 2,570 2,620 50 2 2,630 60 2 Control 7.4 4,120 4,110 -10 0 40 2 4,110 -10 0 50 3 0.9 410 540 130 32 880 470 115 1.9 1,600 1,510 -90 -6 1,920 320 20 3.6 3,040 2,340 -700 -23 2,770 -270 -9 Cold 12.0 4,610 3,840 -770 -17 420 19 4,110 -500 -11 390 39 0.3 850 900 50 6 590 -260 -31 0.8 2,180 2,550 370 17 2,070 -110 -5 1.4 2,770 3,220 450 16 2,820 50 2 Hot 5.1 4,010 4,450 440 11 330 13 4,210 200 5 160 11 324 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-3: Error plot for the Type I - 0.48 mixture, simplified method, E= 40 kJ/mol 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-4: Error plot for the Type I - 0.48 mixture, simplified method, E = 25 kJ/mol 325 Table I-3: Error using NSM function based on simplified method for 30% F mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,180 1,230 50 4 1,230 50 4 1.1 2,200 2,080 -120 -5 2,060 -140 -6 2.0 2,860 2,950 90 3 2,970 110 4 Control 7.0 4,490 4,470 -20 0 70 3 4,470 -20 0 80 4 0.9 670 1,040 370 55 1,320 650 97 1.9 1,990 2,080 90 5 2,420 430 22 3.5 3,090 2,890 -200 -6 3,240 150 5 Cold 12.2 4,990 4,380 -610 -12 320 20 4,590 -400 -8 410 33 0.4 1,350 1,870 520 39 1,380 30 2 0.8 2,860 3,050 190 7 2,520 -340 -12 1.5 3,620 3,820 200 6 3,360 -260 -7 Hot 5.0 4,960 4,920 -40 -1 240 13 4,670 -290 -6 230 7 326 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-5: Error plot for the 30% F mixture, simplified method, E= 40 kJ/mol 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-6: Error plot for the 30% F mixture, simplified method, E = 25 kJ/mol 327 327 Table I-4: Error using NSM function based on simplified method for 20% C mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,220 1,160 -60 -5 1,170 -50 -4 1.0 2,210 2,380 170 8 2,360 150 7 2.0 3,730 3,590 -140 -4 3,610 -120 -3 Control 7.0 5,330 5,360 30 1 100 4 5,350 20 0 90 4 1.2 1,040 1,210 170 16 1,690 650 63 1.8 1,760 2,100 340 19 2,600 840 48 3.4 3,140 3,210 70 2 3,750 610 19 Cold 12.0 5,410 5,130 -280 -5 220 11 5,420 10 0 530 32 0.5 1,930 2,350 420 22 1,710 -220 -11 0.8 3,600 3,690 90 3 3,030 -570 -16 1.4 3,900 4,520 620 16 3,990 90 2 Hot 5.0 5,360 5,810 450 8 400 12 5,530 170 3 260 8 328 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-7: Error plot for the 20% C mixture, simplified method, E= 40 kJ/mol 0 1,750 3,500 5,250 7,000 0 1,750 3,500 5,250 7,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-8: Error plot for the 20% C mixture, simplified method, E = 25 kJ/mol 329 Table I-5: Error using NSM function based on simplified method for 30% C mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 670 710 40 6 730 60 9 1.0 1,940 1,840 -100 -5 1,810 -130 -7 2.0 3,060 3,130 70 2 3,160 100 3 Control 7.2 5,250 5,230 -20 0 60 3 5,230 -20 0 80 5 1.3 570 820 250 44 1,330 760 133 1.9 1,680 1,690 10 1 2,240 560 33 3.5 2,620 2,920 300 11 3,460 840 32 Cold 12.1 4,650 5,230 580 12 290 17 5,450 800 17 740 54 0.4 1,050 1,440 390 37 890 -160 -15 0.8 2,710 3,110 400 15 2,380 -330 -12 1.4 3,900 4,160 260 7 3,510 -390 -10 Hot 5.0 5,870 5,900 30 1 270 15 5,500 -370 -6 310 11 330 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-9: Error plot for the 30% C mixture, simplified method, E= 40 kJ/mol 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-10: Error plot for the 30% C mixture, simplified method, E = 25 kJ/mol 331 Table I-6: Error using NSM function based on simplified method for 30% Slag mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.6 1,430 1,470 40 3 1,480 50 3 1.0 2,730 2,600 -130 -5 2,580 -150 -5 2.0 3,650 3,750 100 3 3,770 120 3 Control 7.1 5,770 5,760 -10 0 70 3 5,750 -20 0 90 3 0.9 610 1,150 540 89 1,550 940 154 1.9 2,050 2,410 360 18 2,910 860 42 3.6 3,400 3,510 110 3 4,040 640 19 Cold 12.2 5,080 5,560 480 9 370 30 5,890 810 16 810 58 0.4 1,500 2,040 540 36 1,530 30 2 0.8 3,130 3,610 480 15 2,990 -140 -4 2.9 4,990 5,770 780 16 5,300 310 6 Hot 4.9 5,240 6,310 1,070 20 720 22 5,950 710 14 300 7 332 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-11: Error plot for the 30% Slag mixture, simplified method, E= 40 kJ/mol 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-12: Error plot for the 30% Slag mixture, simplified method, E = 25 kJ/mol 333 Table I-7: Error using NSM function based on simplified method for 50% Slag mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 760 750 -10 -1 760 0 0 1.0 1,680 1,680 0 0 1,670 -10 -1 2.1 3,060 3,060 0 0 3,070 10 0 Control 7.2 5,770 5,770 0 0 0 0 5,770 0 0 10 0 0.9 330 480 150 45 780 450 136 1.9 1,340 1,530 190 14 1,990 650 49 3.4 2,450 2,640 190 8 3,230 780 32 Cold 12.3 4,960 5,380 420 8 240 19 5,940 980 20 720 59 0.3 610 910 300 49 560 -50 -8 0.8 2,330 2,860 530 23 2,210 -120 -5 1.4 3,240 4,100 860 27 3,350 110 3 Hot 5.0 6,060 6,750 690 11 600 27 6,130 70 1 90 4 334 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-13: Error plot for the 50% Slag mixture, simplified method, E= 40 kJ/mol 0 2,250 4,500 6,750 9,000 0 2,250 4,500 6,750 9,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-14: Error plot for the 50% Slag mixture, simplified method, E = 25 kJ/mol 335 Table I-8: Error using NSM function based on simplified method for Type III - 0.44 mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 730 740 10 1 750 20 3 1.0 3,040 2,970 -70 -2 2,950 -90 -3 2.0 3,820 3,920 100 3 3,950 130 3 Control 7.2 5,100 5,070 -30 -1 50 2 5,060 -40 -1 70 2 1.0 750 1,090 340 45 1,730 980 131 1.8 2,370 2,480 110 5 3,060 690 29 3.4 3,660 3,480 -180 -5 3,980 320 9 Cold 12.1 5,630 4,900 -730 -13 340 17 5,080 -550 -10 640 45 0.3 2,400 2,060 -340 -14 1,360 -1,040 -43 0.7 3,210 3,650 440 14 3,150 -60 -2 1.4 3,740 4,430 690 18 4,080 340 9 Hot 5.3 5,100 5,320 220 4 420 13 5,150 50 1 370 14 336 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-15: Error plot for the Type III - 0.44 mixture, simplified method, E= 40 kJ/mol 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-16: Error plot for the Type III - 0.44 mixture, simplified method, E = 25 kJ/mol 337 Table I-9: Error using NSM function based on simplified method for 70/20/10 - 0.37 mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.5 1,300 1,320 20 2 1,330 30 2 1.0 3,650 3,590 -60 -2 3,560 -90 -2 2.1 5,240 5,290 50 1 5,330 90 2 Control 7.0 7,670 7,650 -20 0 40 1 7,640 -30 0 60 2 0.9 1,040 1,560 520 50 2,030 990 95 1.8 3,060 3,430 370 12 4,040 980 32 3.5 4,650 4,880 230 5 5,540 890 19 Cold 12.2 7,000 7,460 460 7 400 14 7,820 820 12 920 25 0.3 1,640 2,190 550 34 1,330 -310 -19 0.7 3,540 5,070 1,530 43 4,130 590 17 1.4 5,020 6,540 1,520 30 5,730 710 14 Hot 4.9 7,390 8,350 960 13 1140 23 7,910 520 7 530 13 338 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-17: Error plot for the 70/20/10 - 0.37 mixture, simplified method, E= 40 kJ/mol 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-18: Error plot for the 70/20/10 - 0.37 mixture, simplified method, E = 25 kJ/mol 339 Table I-10: Error using NSM function based on simplified method for the 70/20/10 - 0.44 mixture E = 40,000 J/mol E = 25,000 J/mol Batch ID Concrete Age (days) Compressive Strength Test Result (psi) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) Estimated Strength (psi) Error (psi) EoE (%) AAE (psi) Abs. EoE (%) 0.4 510 540 30 6 550 40 8 1.0 1,770 1,690 -80 -5 1,670 -100 -6 2.0 2,610 2,670 60 2 2,690 80 3 Control 7.2 4,680 4,670 -10 0 50 3 4,660 -20 0 60 4 1.0 370 790 420 114 1,090 720 195 1.8 1,450 1,470 20 1 1,850 400 28 3.4 2,560 2,370 -190 -7 2,850 290 11 Cold 12.1 4,510 4,480 -30 -1 170 31 4,820 310 7 430 60 0.3 980 1,220 240 24 790 -190 -19 0.8 1,860 2,590 730 39 2,010 150 8 1.5 2,720 3,690 970 36 3,090 370 14 Hot 5.0 4,500 5,390 890 20 710 30 4,920 420 9 280 13 340 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-19: Error plot for the 70/20/10 - 0.44 mixture, simplified method, E= 40 kJ/mol 0 2,000 4,000 6,000 8,000 0 2,000 4,000 6,000 8,000 Measured Strength (psi) Estimated Strength (psi) Control Data Cold Data Hot Data 10% Error 20% Error Figure I-20: Error plot for the 70/20/10 - 0.44 mixture, simplified method, E = 25 kJ/mol