DEVELOPING SIMPLE LAB TEST TO EVALUATE HMA RESISTANCE TO MOISTURE, RUTTING, THERMAL CRACKING DISTRESS

A Dissertation Presented to The Graduate Faculty of The University of Akron

In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

Feng Zhu

March, 2008

DEVELOPING SIMPLE LAB TEST TO EVALUATE HMA RESISTANCE TO MOISTURE, RUTTING, THERMAL CRACKING DISTRESS

Feng Zhu

Dissertation

Approved: Advisor Dr. Robert Y. Liang Committee Member Dr. Ala R. Abbas Committee Member Dr. Daren J. Zywicki Committee Member Dr. Yueh-Jaw Lin Committee Member Dr. Kevin L Kreider

Accepted: Department Chair Dr. Wieslaw K. Binienda Dean of the College Dr. George K. Haritos Dean of the Graduate School Dr. George R. Newkome Date

ii

ABSTRACT

The pavement distresses including stripping, rutting, and low temperature cracking have been known to greatly decrease the service life of pavement. The simple laboratory tests are used to measure HMAs (Hot Mix Asphalt) fundamental engineering properties that can be linked back to the pavement distress-prediction models. It is envisioned that these simple laboratory tests would play a key role in the quality control of HMA mixtures. To minimize premature pavement distresses, it is imperative to investigate the relationship between the simple laboratory tests and pavement distresses. Moisture induced damage in asphalt concrete pavement, better known as stripping, is a primary cause of distresses in the Hot Mix Asphalt layer. AASHTO T 283 has historically been used to detect moisture damage of asphalt mixes. This method is established for the Marshall mix design process. However, the current hot mix asphalt design calls for the use of Superpave mix design procedure using the Superpave Gyratory Compactor. The differences in the mix design methods will most likely introduce significant differences in stripping test results. Therefore, there is a need to improve the AASHTO T 283 test procedure and to develop the new criterion for the Superpave HMA specimens. A structured laboratory program test is conducted in this research to study the effects of the various factors on the HMA specimens susceptibility to moisture damage. The experimental test program is divided into two parts: one part is a complete factorial iii

experimental program for two aggregate sources with virgin asphalt binder, the second part is a partial factorial experimental program for the other two aggregate sources with polymer modified asphalt binders. The complete factorial experiment is used to identify and evaluate the contributions of the individual factors, along with all the possible interactions among the various factors. The partial factorial experiment, on the other hand, is used for validating findings from complete factorial experimental results. In the data analysis, the effects of different factors on dry tensile strength, freeze-thaw conditioned tensile strength, and tensile strength ratio (TSR) are investigated in this research. The effects of all the factors investigated in this research are summarized in this research. The recommendations for the proposed stripping test procedure for Superpave HMA are given finally. Rutting and low temperature cracking are also primary failure modes of Hot Mix Asphalt pavements. In the mid-1980s, polymer-modified asphalt concrete mixture has been widely used to minimize rutting and low temperature cracking failures of flexible pavement. In this research, three polymer modifiers are used in the laboratory test. The influences of the modifiers to rutting resistance and low temperature cracking resistance are investigated and regression procedures are performed in an attempt to develop mathematical relationships between pavement distresses properties and HMAs physical properties and strength properties obtained from the simple laboratory tests.

iv

ACKNOWLEDGEMENTS

First and foremost, I would like to express my sincere appreciation to my advisor, Professor Robert Y. Liang, for his valuable guidance, continuous support and encouragement throughout this study. He encouraged me to not only grow as an engineer but also as an independent thinker. The useful discussions and guidance regarding the distress of pavement from Dr. Ala R. Abbas are really appreciated. I would also like to thank my committee members, Dr. Daren J. Zywicki, Dr. Yueh-Jaw Lin, Dr. Kevin L Kreider, for reviewing my work and helpful recommendations. I would like to thank my fellow graduate students, Dr. Ke Yang, Dr. Luo Yang, Songquan Wang, Ruifeng Wang, Yan Liu, Jiliang Li, Abdulla A Sharo, Ehab S Shatnawi, Madhar Mohammad Ta'amneh, Wassel M Al Bodour, Mohammad A Khasawneh, and Mohd M. Yamin. The sincere friendship and support from my friends in Akron always give me energy and impetus to finish this dissertation. Finally, and most importantly, I would like to thank my wife Qiong Zhou. Her support, encouragement, quiet patience and unwavering love were undeniably the bedrock upon which the past four years of my life have been built. I thank my parents, Jiarong Zhu, Keyu Zhang, Jianxin Zhou, and Faqing Dai for their unending encouragement, support and love. v

TABLE OF CONTENTS Page LIST OF TABLES......................................................................................................... xi LIST OF FIGURES ................................................................................................... xviii CHAPTER I. INTRODUCTION ....................................................................................................... 1 1.1 Problem Statement ............................................................................................ 1 1.1.1 Moisture Damage................................................................................... 1 1.1.2 Rutting and Low Temperature Cracking ............................................... 5 1.2 Objectives of the Study..................................................................................... 7 1.2.1 Moisture Damage................................................................................... 7 1.2.2 Rutting and Low Temperature Cracking ............................................... 8 1.3 Organization of Dissertation ............................................................................. 8 II. LITERATURE REVIEW......................................................................................... 10 2.1 Literature Review of Moisture Damage ......................................................... 10 2.1.1 AASHTO T-283 Test Method ............................................................. 11 2.1.2 Efforts to Develop Alternative Test Methods...................................... 16

vi

2.1.3 Previous Experience in Modifying the AASHTO T 283..................... 20 2.1.4 Summary .............................................................................................. 23 2.2 Literature Review of Rutting .......................................................................... 26 2.2.1 Previous Findings Based on the Unmodified Asphalt ......................... 27 2.2.2 Previous Findings Based on the Modified Asphalt.............................. 29 2.3 Literature Review of Low Temperature Cracking .......................................... 29 III. FACTOTS AFFECTING WATER STRIPPING TEST RESULTS ...................... 34 3.1 Introduction..................................................................................................... 34 3.2 Experimental Program .................................................................................... 35 3.2.1 Introduction of Research Approach ..................................................... 35 3.2.2 Materials Tested................................................................................... 36 3.3 Data Analysis and Discussion......................................................................... 40 3.3.1 Dry Tensile Strength ............................................................................ 40 3.3.2 Freeze-Thaw Tensile Strength ............................................................. 53 3.3.3 Dry Tensile Strength versus Freeze-Thaw Tensile Strength ............... 71 3.3.4 Level of Saturation............................................................................... 79 3.3.5 Water Sensitivity................................................................................. 86 3.3.6 Tensile Strength Ratio.......................................................................... 90 3.4 Summary and Conclusions ............................................................................. 95 IV. DEVELOPMENT OF STRIPPING TEST PROCEDURURE FOR SUPERPAVE SPECIMENS......................................................................................... 97 4.1 Introduction..................................................................................................... 97 vii

4.2 Analysis of Tensile Strength Ratio.................................................................. 97 4.3 Partial Factorial Test Program....................................................................... 104 4.4 The Partial Factorial Test Results and Validation of Proposed Procedure.... 107 4.5 Conclusions and Recommendations ............................................................. 109 V. HMA RUT RESISTANCE CHARACTERIZED BY SIMPLE MATERIAL PROPERTIES ............................................................................................................. 111 5.1 Introduction and Objectives of the Chapter ...................................................111 5.2 Research Plan and Experimental Data...........................................................111 5.2.1 Research Approach ............................................................................ 111 5.2.2 Materials ............................................................................................ 113 5.2.3 Mixture Designs and Physical Properties .......................................... 114 5.2.4 Asphalt Mixture Test and Mechanical Strength Properties ............... 115 5.2.5 Rutting Test Data from APA Test ..................................................... 117 5.3 Analysis of Test Results .................................................................................118 5.3.1 Single Parameter Linear Regression Analysis................................... 118 5.3.2 Multiple Parameter Linear Regression Analysis ............................... 125 5.4 Conclusions................................................................................................... 128 VI. RELATIOSHIP INVESTIGATION BETWEEN LOW TEMPERATURE CRACKING AND SIMPLE LABORATORY TEST................................................ 130 6.1 Introduction................................................................................................... 130 6.2 Experimental Tests........................................................................................ 131

viii

6.3 Analysis of Test Results ................................................................................ 135 6.4 Conclusions and Recommendations ............................................................. 143 VII. CONCLUSIONS AND RECOMMENDATIONS.............................................. 144 7.1 Summary of Moisture Damage Conclusions and Recommendations........... 144 7.1.1 Summary of the Moisture Damage Work Completed ....................... 144 7.1.2 Conclusions and Recommendations .................................................. 146 7.1.3 Implementation Recommendations ................................................... 147 7.1.4 Secondary Observations of Experimental Results............................. 148 7.1.5 Recommendations for Future Study .................................................. 151 7.2 Summary of the Rutting and Thermal Cracking Conclusions and Recommendations............................................................................................... 151 REFERENCES ........................................................................................................... 154 APPENDICES ............................................................................................................ 160 APPENDIX A. TEST RESULTS OF THE COMPLETE FACTORIAL EXPERIMENTAL PROGRAM ......................................................................... 161 A.1 Complete Factorial Experimental Program Test Results for Honey Creek Limestone ......................................................................................... 161 A.2 Complete Factorial Experimental Program Test Results for Martin Marietta Gravel ........................................................................................... 170 APPENDIX B. OPTIMUM ASPHALT CONTENT MIX DESIGN ................ 179 B.1 Honey Creek Limestone...................................................................... 180 B.2 Martin Marietta Gravel........................................................................ 200 ix

B.3 Traprock Source .................................................................................. 221 B.4 Stocker Sand Gravel............................................................................ 229

LIST OF TABLES Table Page

2-1 Comparison of Standard Water Sensitivity Tests....................................................... 24 2-2 Summary of State Agencies Experience in Modifying the AASHTO T 283 in different U.S. Departments of Transportation................................... 25 3-1 Superpave Asphalt Binder Specification for the PG 64-22........................................ 39 3-2 Statistical Comparison of 100-mm Diameter Marshall Compacted Samples and 100-mm Diameter Superpave Gyratory Compacted Samples Dry Tensile Strength) ....................................................................... 41 3-3 Statistical Comparison of 100-mm Diameter Marshall Compacted Samples and 150-mm Diameter Superpave Gyratory Compacted Samples (Dry Tensile Strength)...................................................................... 41 3-4 Statistical Comparison of 100-mm and 150-mm Diameter Superpave Gyratory Compacted Samples (Dry Tensile Strength) ................................... 41 3-5 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Loose Mix Aging (Honey Creek Limestone) ............................ 42 3-6 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Loose Mix Aging (Martin Marietta Gravel) .............................. 43 3-7 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Honey Creek Limestone) .................... 43 3-8 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Martin Marietta Gravel)...................... 44 3-9 ANOVA Analysis of the Dry Tensile Strength for Honey Creek Limestone (Marshall 100-mm specimen) ....................................................... 45 3-10 ANOVA Analysis of the Dry Tensile Strength for Honey Creek Limestone (Gyratory 100-mm specimen)....................................................... 45 xi

3-11 ANOVA Analysis of the Dry Tensile Strength for Honey Creek Limestone (Gyratory 150-mm specimen)....................................................... 45 3-12 Summarizing Comparison ANOVA Analysis for the Dry Tensile Strength for Honey Creek Limestone ............................................................. 46 3-13 ANOVA Analysis of the Dry Tensile Strength for Martin Marietta Gravel (Marshall 100-mm specimen) ............................................................. 47 3-14 ANOVA Analysis of the Dry Tensile Strength for Martin Marietta Gravel (Gyratory 100-mm specimen)............................................................. 47 3-15 ANOVA Analysis of the Dry Tensile Strength for Martin Marietta Gravel (Gyratory 150-mm specimen)............................................................. 47 3-16 Summarizing Comparison ANOVA Analysis for the Dry Tensile Strength for Martin Marietta Gravel ............................................................... 48 3-17 Results Summary of Dry Tensile Strength for the Limestone.................................. 49 3-18 Results Summary of Dry Tensile Strength for the Gravel........................................ 50 3-19 Class level Information for Dry Tensile Strength .................................................... 50 3-20 ANOVA Analysis of Dry Tensile Strength.............................................................. 51 3-21 Statistical Comparison of 100-mm Diameter Marshall Compacted Samples and 100-mm Diameter Superpave Gyratory Compacted Samples (F-T Tensile Strength) ...................................................................... 54 3-22 Comparison of 100-mm Diameter Marshall Compacted Samples and 150-mm Diameter Superpave Gyratory Compacted Samples (F-T Tensile Strength) ...................................................................... 54 3-23 Statistical Comparison of 100-mm and 150-mm Diameter Superpave Gyratory Compacted Samples (F-T Tensile Strength) ................................... 54 3-24 Statistical Comparison of F-T Tensile Strength for Mixtures Subjected to Loose Mix Aging (Honey Creek Limestone) ............................................. 56 3-25 Statistical Comparison of F-T Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Honey Creek Limestone) .................... 57 3-26 Statistical Comparison of F-T Tensile Strength for Mixtures Subjected to Loose Mix Aging (Martin Marietta Gravel) .............................................. 58 xii

3-27 Statistical Comparison of F-T Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Martin Marietta Gravel)...................... 59 3-28 ANOVA Analysis of the Complete Factorial Experiment for the Freeze-Thaw Tensile Strength for Honey Creek Limestone (Marshall 100mm specimen) ........................................................................ 61 3-29 ANOVA Analysis of the Complete Factorial Experiment for the Freeze-Thaw Tensile Strength for Honey Creek Limestone (Gyratory 100mm specimen) ........................................................................ 61 3-30 ANOVA Analysis of the Complete Factorial Experiment for the Freeze-Thaw Tensile Strength for Honey Creek Limestone (Gyratory 150mm specimen) ........................................................................ 61 3-31 Comparison ANOVA Analysis for the F-T Tensile Strength for Honey Creek Limestone ................................................................................. 62 3-32 ANOVA Analysis of the F-T Tensile Strength for Martin Marietta Gravel (Marshall 100-mm specimen) ............................................................. 63 3-33 ANOVA Analysis of the F-T Tensile Strength for Martin Marietta Gravel (Gyratory 100-mm specimen)............................................................. 64 3-34 ANOVA Analysis of the F-T Tensile Strength for Martin Marietta Gravel (Gyratory 150-mm specimen)............................................................. 64 3-35 Comparison ANOVA Analysis for the F-T Tensile Strength for Martin Marietta Gravel ............................................................................................... 65 3-36 Results Summary of F-T Tensile Strength for the Limestone.................................. 67 3-37 Results Summary of F-T Tensile Strength for the Gravel........................................ 68 3-38 ANOVA Analysis of F-T Tensile Strength .............................................................. 69 3-39 Statistical Comparison of Dry Tensile Strength and Freeze-Thaw Tensile Strength (Honey Creek Limestone) ................................................... 73 3-40 Statistical Comparison of Dry Tensile Strength and Freeze-Thaw Tensile Strength (Martin Marietta Gravel) ..................................................... 74 3-41 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 70% vs. 50% (Honey Creek Limestone) ....... 80

xiii

3-42 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 50% (Honey Creek Limestone) ....... 81 3-43 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 70% (Honey Creek Limestone) ....... 82 3-44 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 70% vs. 50% (Martin Marietta Gravel) ......... 83 3-45 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 50% (Martin Marietta Gravel) ......... 84 3-46 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 70% (Martin Marietta Gravel) ......... 85 3-47 Acceptable Mixtures................................................................................................. 87 3-48 Tensile Strength Ratio for Freeze-Thaw Conditioning (Honey Greek Limestone)............................................................................... 88 3-49 Tensile Strength Ratio for Freeze-Thaw Conditioning (Martin Marietta Gravel)................................................................................. 89 3-50 Comparison of TSR of M100, G100 and G150 at Different Conditions (Honey Greek Limestone)............................................................................... 91 3-51 Comparison of TSR of M100, G100 and G150 at Different Conditions (Martin Marietta Gravel)................................................................................. 92 3-52 ANOVA Analysis of the Tensile Strength Ratio for Limestone .............................. 93 3-53 ANOVA Analysis of the Tensile Strength Ratio for Gravel .................................... 93 4-1 Summary of Regression between TSR values and the Degree of Saturation........... 104 4-2 Summary of the Suggested Test Procedure for Gyratory 150-mm (G150) Specimen .......................................................................................... 105 4-3 Comparison of TSR of M100 at Standard Procedure and G150 at Proposed Procedure ...................................................................................... 108 5-1 Optimum Properties of Marshall Design for All Mixture ........................................ 114 5-2 Indirect Tensile Strength Test Results and Unconfined Compression Test Results for All Mixture ......................................................................... 116 xiv

5-3 Asphalt Pavement Analyzer Test Results for All Mixture ....................................... 117 5-4 Summary of Regression Analysis Result between Rut Depth and Mixture Properties.................................................................................. 120 5-5 Linear Regression Analysis Results for Model Summary........................................ 126 5-5 Significance of the Model......................................................................................... 127 6-1 TSRST Test Results for All Mixture........................................................................ 137 6-2 Indirect Tensile Strength, Resilient Modulus, and Dynamic Modulus Test Results for All Mixture .......................................................... 138 A-1 Test Results for Marshall 100 Specimen with No Loose Mix Aging (Honey Creek Limestone)............................................................................. 161 A-2 Test Results for Marshall 100 Specimen with 4 Hrs Loose Mix Aging (Honey Creek Limestone)............................................................................. 162 A-3 Test Results for Marshall 100 Specimen with 16 Hrs Loose Mix Aging (Honey Creek Limestone)............................................................................. 163 A-4 Test Results for Gyratory 100 Specimen with No Loose Mix Aging (Honey Creek Limestone)............................................................................. 164 A-5 Test Results for Gyratory 100 Specimen with 4 Hrs Loose Mix Aging (Honey Creek Limestone)............................................................................. 165 A-6 Test Results for Gyratory 100 Specimen with 16 Hrs Loose Mix Aging (Honey Creek Limestone)............................................................................. 166 A-7 Test Results for Gyratory 150 Specimen with No Loose Mix Aging (Honey Creek Limestone)............................................................................. 167 A-8 Test Results for Gyratory 150 Specimen with 4 Hrs Loose Mix Aging (Honey Creek Limestone)............................................................................. 168 A-9 Test Results for Gyratory 150 Specimen with 16 Hrs Loose Mix Aging (Martin Marietta Gravel)............................................................................... 169 A-10 Test Results for Marshall 100 Specimen with No Loose Mix Aging (Martin Marietta Gravel)............................................................................... 170

xv

A-11 Test Results for Marshall 100 Specimen with 4 Hrs Loose Mix Aging (Martin Marietta Gravel)............................................................................... 171 A-12 Test Results for Marshall 100 Specimen with 16 Hrs Loose Mix Aging (Martin Marietta Gravel)............................................................................... 172 A-13 Test Results for Gyratory 100 Specimen with No Loose Mix Aging (Martin Marietta Gravel)............................................................................... 173 A-14 Test Results for Gyratory 100 Specimen with 4 Hrs Loose Mix Aging (Martin Marietta Gravel)............................................................................... 174 A-15 Test Results for Gyratory 100 Specimen with 16 Hrs Loose Mix Aging (Martin Marietta Gravel)............................................................................... 175 A-16 Test Results for Gyratory 150 Specimen with No Loose Mix Aging (Martin Marietta Gravel)............................................................................... 176 A-17 Test Results for Gyratory 150 Specimen with 4 Hrs Loose Mix Aging (Martin Marietta Gravel)............................................................................... 177 A-18 Test Results for Gyratory 150 Specimen with 16 Hrs Loose Mix Aging (Martin Marietta Gravel)............................................................................... 178 B-1 Marshall Mix Design for Honey Creek Limestone.................................................. 180 B-2 Compaction Data for 5% AC (G4 for Honey Creek Limestone)............................. 184 B-3 Compaction Data for 5.5% AC (G4 for Honey Creek Limestone).......................... 185 B-4 Compaction Data for 6% AC (G4 for Honey Creek Limestone)............................. 186 B-5 Compaction Data for 6.5% AC (G4 for Honey Creek Limestone).......................... 187 B-6 Summary of Compaction Data (G4 for Honey Creek Limestone) ......................... 188 B-7 Summary of Design Parameters (G4 for Honey Creek Limestone) ........................ 188 B-8 Compaction Data for 4% AC (G6 for Honey Creek Limestone)............................. 192 B-9 Compaction Data for 4.5 % AC (G6 for Honey Creek Limestone)......................... 193 B-10 Compaction Data for 5% AC (G6 for Honey Creek Limestone)........................... 194 B-11 Compaction Data for 5.5% AC (G6 for Honey Creek Limestone)........................ 195 xvi

B-12 Summary of Compaction Data (G6 for Honey Creek Limestone) ....................... 196 B-13 Summary of Design Parameters (G6 for Honey Creek Limestone) ...................... 196 B-14 Marshall Mix Design for Martin Marietta Gravel ................................................. 200 B-15 Compaction Data for 5% AC (G4 for Martin Marietta Gravel) ............................ 204 B-16 Compaction Data for 5% AC (G4 for Martin Marietta Gravel) ............................ 205 B-17 Compaction Data for 5% AC (G4 for Martin Marietta Gravel) ............................ 206 B-18 Compaction Data for 5% AC (G4 for Martin Marietta Gravel) ........................... 207 B-19 Summary of Compaction Data (G4 for Martin Marietta Gravel) .......................... 208 B-20 Summary of Design Parameters (G4 for Martin Marietta Gravel) ........................ 208 B-21 Compaction Data for 4% AC (G6 for Martin Marietta Gravel) ............................ 212 B-22 Compaction Data for 4.5% AC (G6 for Martin Marietta Gravel) ......................... 213 B-23 Compaction Data for 5% AC (G6 for Martin Marietta Gravel) ............................ 214 B-24 Compaction Data for 5.5% AC (G6 for Martin Marietta Gravel) ......................... 215 B-25 Compaction Data for 6% AC (G6 for Martin Marietta Gravel) ............................ 216 B-26 Summary of Compaction Data (G6 for Martin Marietta Gravel) .......................... 217 B-27 Summary of Design Parameters (G6 for Martin Marietta Gravel) ........................ 217 B-28 Marshall Mix Design for Ontario Traprock........................................................... 221 B-29 Summary of Design Parameters (G6 for Ontario Traprock) ................................. 225

xvii

LIST OF FIGURES Figure Page

1-1 HMA at the Time of Mixing (McCann et al., 2001) .................................................... 4 2-1 The Superpave Gyratory Compactor.......................................................................... 13 2-2 The Marshall Impact Compactor................................................................................ 14 2-3 The Saturation Chamber............................................................................................. 15 2-4 Water Bath Container ................................................................................................. 15 2-5 MTS Machine ............................................................................................................. 16 3-1 Complete Factorial Experimental Test Program ........................................................ 38 3-2 ODOT Requirements for Heavy Traffic (Type-1H)................................................... 39 3-3 Distribution of Dry Tensile Strength for Honey Creek Limestone ............................ 52 3-4 Distribution of Dry Tensile Strength for Martin Marietta Gravel ............................. 52 3-5 Distribution of Freeze-Thaw Conditioned Tensile Strength for Limestone............... 70 3-6 Distribution of Freeze-Thaw Conditioned Tensile Strength for Gravel..................... 70 3-7 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Honey Creek Limestone (Marshall 100-mm Diameter Specimen)............. 76 3-8 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Honey Creek Limestone (Gyratory 100-mm Diameter Specimen) ............ 76 3-9 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Honey Creek Limestone (Gyratory 150-mm Diameter Specimen) ............ 77 3-10 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Martin Marietta Gravel (Marshall 100-mm Diameter Specimen) .............. 77 xviii

3-11 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Martin Marietta Gravel (Gyratory 100-mm Diameter Specimen) .............. 78 3-12 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Martin Marietta Gravel (Gyratory 150-mm Diameter Specimen) .............. 78 3-13 Relationship of the F-T Tensile Strength Versus. Saturation Level......................... 86 3-14 TSR distributions of M100, G100 and G150 at different conditions (Honey Creek Limestone).............................................................................................. 94 3-15 TSR distributions of M100, G100 and G150 at different conditions (Martin Marietta Gravel)................................................................................................ 94 4-1 TSR Regression Simulations for Honey Creek Limestone (M100 Vs. G100) ............................................................................................ 100 4-2 TSR Regression Simulations for Honey Creek Limestone (M100 Vs. G150) ............................................................................................ 100 4-3 TSR Regression Simulations for Martin Marietta Gravel (M100 Vs. G100) ............................................................................................ 101 4-4 TSR Regression Simulations for Martin Gravel (M100 Vs. G150) ............................................................................................ 101 4-5 Comparisons of TSR Values of M100 at Standard Loose Mix Aging Condition with G150 at different Loose Mix Aging Condition (Limestone).................. 102 4-6 Comparisons of TSR Values of M100 at Standard Loose Mix Aging Condition with G150 at different Loose Mix Aging Condition (Gravel)........................ 102 4-7 Relationship of TSR and Saturation Level for Honey Creek Limestone ................. 103 4-8 TSR Values of M100 and G150 with Different Saturation Level............................ 103 4-9 Partial Factorial Experimental Test Program ........................................................... 106 5-1 Framework of the Research Approach ..................................................................... 112 5-2 ODOT Requirements for Heavy Traffic (Type-1H)................................................. 113 5-3 Rutting Depth Vs. Number of Cycles for 2% and 3% Modifiers............................. 120 5-4 Rutting Depth Vs. Number of Cycles for 4% and 5% Modifiers............................. 121 xix

5-5 Rutting Depth Vs. Modifiers Content After 8000 cycles ......................................... 121 5-6 Regression Simulations for Indirect Tensile Strength.............................................. 123 5-7 Regression Simulations for Unconfined Compressive Strength .............................. 123 5-8 Indirect Tensile Strength Vs. Modified Content for all Mixture.............................. 124 5-9 Compressive Strength Vs. Modified Content for all Mixture .................................. 124 5-10 Predicted Vs. Measured Rutting Potential............................................................. 127 6-1 Schematic of TSRST apparatus................................................................................ 133 6-2 Schematic of Specimen Alignment Stand ................................................................ 133 6-3 Fracture Temperature Vs. Modified Content in TSRST .......................................... 139 6-4 Fracture Tensile Strength Vs. Modified Content in TSRST .................................... 139 6-5 Regression between TSRST Fracture Strength and Indirect Tensile Strength......... 140 6-6 Regression between TSRST Fracture Strength and Resilient Modulus ................... 140 6-7 Regression between TSRST Fracture Strength and Dynamic Modulus................... 141 6-8 Regression between TSRST Fracture Temperature and Indirect Tensile Strength .............................................................................................. 141 6-9 Regression between TSRST Fracture Temperature and Resilient Modulus ............ 142 6-10 Regression between TSRST Fracture Temperature and Dynamic Modulus.......... 142 B-1 Marshall Mix Design (AV Vs. AC for Honey Creek Limestone) ........................... 180 B-2 Marshall Mix Design (VMA Vs. AC for Honey Creek Limestone)........................ 181 B-3 Marshall Mix Design (Stability Vs. AC for Honey Creek Limestone) ................... 181 B-4 Marshall Mix Design (Flow Vs. AC for Honey Creek Limestone)......................... 182 B-5 Marshall Mix Design (VFA Vs. AC for Honey Creek Limestone)......................... 182 B-6 Marshall Mix Design (Unit WGT Vs. AC for Honey Creek Limestone)................ 183

xx

B-7 Gyratory 4 Mix Design (Gmm% Vs. Gyration No. for Honey Creek Limestone).. 189 B-8 Gyratory 4 Mix Design (Gmm% at Initial Vs. AC for Honey Creek Limestone)... 189 B-9 Gyratory 4 Mix Design (Gmm% at Max Vs. AC for Honey Creek Limestone) ..... 190 B-10 Gyratory 4 Mix Design (Air Void Vs. AC for Honey Creek Limestone) ............. 190 B-11 Gyratory 4 Mix Design (VMA Vs. AC for Honey Creek Limestone) .................. 191 B-12 Gyratory 4 Mix Design (VFA Vs. AC for Honey Creek Limestone)................... 191 B-13 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Honey Creek Limestone) 197 B-14 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Honey Creek Limestone). 197 B-15 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Honey Creek Limestone) ... 198 B-16 Gyratory 6 Mix Design (Air Void Vs. AC for Honey Creek Limestone) ............. 198 B-17 Gyratory 6 Mix Design (VMA Vs. AC for Honey Creek Limestone) .................. 199 B-18 Gyratory 6 Mix Design (VFA Vs. AC for Honey Creek Limestone).................... 199 B-19 Marshall Mix Design (AV Vs. AC for Martin Marietta Gravel)........................... 201 B-20 Marshall Mix Design (VMA Vs. AC for Martin Marietta Gravel) ....................... 201 B-21 Marshall Mix Design (Stability Vs. AC for Martin Marietta Gravel) ................... 202 B-22 Marshall Mix Design (Flow Vs. AC for Martin Marietta Gravel) ........................ 202 B-23 Marshall Mix Design (VFA Vs. AC for Martin Marietta Gravel)......................... 203 B-24 Marshall Mix Design (VMA Vs. AC for Martin Marietta Gravel) ....................... 203 B-25 Gyratory 4 Mix Design (Gmm% Vs. Gyration No. for Martin Marietta Gravel) . 209 B-26 Gyratory 4 Mix Design (Gmm% at Initial Vs. AC for Martin Marietta Gravel)... 209 B-27 Gyratory 4 Mix Design (Gmm% at Max Vs. AC for Martin Marietta Gravel)..... 210 B-28 Gyratory 4 Mix Design (Air Void Vs. AC for Martin Marietta Gravel) ............... 210 B-29 Gyratory 4 Mix Design (VMA Vs. AC for Martin Marietta Gravel) .................... 211 xxi

B-30 Gyratory 4 Mix Design (VFA Vs. AC for Martin Marietta Gravel) ..................... 211 B-31 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Martin Marietta Gravel) . 218 B-32 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Martin Marietta Gravel)... 218 B-33 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Martin Marietta Gravel)..... 219 B-34 Gyratory 6 Mix Design (Air Void Vs. AC for Martin Marietta Gravel) ............... 219 B-35 Gyratory 6 Mix Design (VMA Vs. AC for Martin Marietta Gravel) .................... 220 B-36 Gyratory 6 Mix Design (VFA Vs. AC for Martin Marietta Gravel) ..................... 220 B-37 Marshall Mix Design (AV Vs. AC for Ontario Traprock) .................................... 222 B-38 Marshall Mix Design (VMA Vs. AC for Ontario Traprock)................................. 222 B-39 Marshall Mix Design (Stability Vs. AC for Ontario Traprock)............................. 223 B-40 Marshall Mix Design (Flow Vs. AC for Ontario Traprock).................................. 223 B-41 Marshall Mix Design (VFA Vs. AC for Ontario Traprock) .................................. 224 B-42 Marshall Mix Design (Unit WGT Vs. AC for Ontario Traprock)......................... 224 B-43 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Ontario Traprock)........... 225 B-44 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Ontario Traprock) ............ 226 B-45 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Ontario Traprock) .............. 226 B-46 Gyratory 6 Mix Design (Air Void Vs. AC for Ontario Traprock)......................... 227 B-47 Gyratory 6 Mix Design (VMA Vs. AC for Ontario Traprock).............................. 227 B-48 Gyratory 6 Mix Design (Air Void Vs. AC for Ontario Traprock)......................... 228 B-49 Marshall Mix Design (AV Vs. AC for Stocker Sand Gravel) ............................... 230 B-50 Marshall Mix Design (VMA Vs. AC for Stocker Sand Gravel)............................ 230 B-51 Marshall Mix Design (Stability Vs. AC for Stocker Sand Gravel) ....................... 231

xxii

B-52 Marshall Mix Design (Flow Vs. AC for Stocker Sand Gravel)............................. 231 B-53 Marshall Mix Design (VFA Vs. AC for Stocker Sand Gravel)............................. 232 B-54 Marshall Mix Design (Unit WGT Vs. AC for Stocker Sand Gravel).................... 232 B-55 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Stocker Sand Gravel)...... 234 B-56 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Stocker Sand Gravel)....... 235 B-57 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Stocker Sand Gravel) ......... 235 B-58 Gyratory 6 Mix Design (Air Void Vs. AC for Stocker Sand Gravel) ................... 236 B-59 Gyratory 6 Mix Design (VMA Vs. AC for Stocker Sand Gravel) ........................ 236 B-60 Gyratory 6 Mix Design (VFA Vs. AC for Stocker Sand Gravel).......................... 237

xxiii

CHAPTER I INTRODUCTION

1.1 Problem Statement Pavement distresses including moisture damage, rutting and thermal cracking are primary distresses in the asphalt pavement layers. 1.1.1 Moisture Damage Moisture damage in asphalt concrete pavements is a primary cause of distresses in the asphalt pavement layers. The existence of water in asphalt pavement is often one of the major factors affecting the durability of HMA. The water induced damage in HMA layers may be associated with two mechanisms: loss of adhesion and/loss of cohesion. In the first mechanism, the water gets between the asphalt and aggregate and strips the asphalt film away, leaving aggregate without asphalt film coverage, as illustrated in Figure1-1 and Figure1-2. This is because the aggregates have a greater kinship for water than asphalt binder. The second mechanism includes the interaction of water with the asphalt cement that reduces the cohesion within the asphalt cement. The water sensitivity test methods listed below are national standards and are used by public agencies (AASHTO and ASTM):

 AASHTO T283, Resistance of Compacted Bituminous Mixture to Moisture Induced Damage ASTM D4867, Effect of Moisture on Asphalt Concrete Paving Mixtures AASHTO T165/ASTM D1075, Effect of Water on Compressive Strength of Compacted Bituminous Mixtures ASTM D3625, Effect of Water on Bituminous-Coated Aggregate Using Boiling Water. AASHTO T283 is based on research performed by R. P. Lottman under NCHRP Project 4-08(03) and subsequent research performed by D. G. Tunnicliff and R. E. Root under NCHRP Project 10-17. The AASHTO method indicates that it is suitable for testing samples prepared as part of the mixture design process (i.e., laboratory-mixedlaboratory-compacted), as part of the plant control process (i.e., field-mixedlaboratory-compacted) and for cores taken from the roadway (i.e., field-mixedfield-compacted). The AASHTO procedure ages the mixed, loose HMA for 16 hr at 60 C. After compaction to an air-void content of 7 percent 1 percent, the samples are extruded from the compaction mold and allowed to age 24 hours at room temperature. The samples are then placed under water, and a vacuum is used to saturate the samples to a degree of saturation level between 55-80 percent (AASHTO T283-99) or 70-80 percent (AASHTO T283-03). A freeze cycle (16 hr at -18 C) and a thaw-soak cycle (24 hr at 60 C) are used to condition the sample prior to indirect tension testing at 25 C. The most widely used method for determining HMA moisture resistance is the AASHTO Standard Method of Test T 283, Resistance of Compacted Bituminous Mixture to Moisture Induced Damage. It is established for the Marshall mix process that 2

uses 4 inch specimens. On the other hand, the Superpave mix design process is being conducted using 6 inch diameter specimens and with a totally different compaction device, Superpave Gyratory Compactor (SGC), for compacting and preparing the specimens. The differences in the mix design methods will most likely introduce significant differences in the moisture resistance test results for the same materials. Currently, SHRP recommended the use of AASHTO T 283 to evaluate the water sensitivity of HMA within the Superpave volumetric mixture design system. Recently, NCHRP 444 (2000) and FHWA/IN/JTRP-97/13 (1997) studies called the public agencies to conduct their own experiments, using their aggregate and asphalt binder sources, before making any modifications to the AASHTO T 283. In 2004 TRB Annual Meeting, Hicks et al. (2004) summarized the road map for mitigating national moisture sensitivity concerns in hot mix pavements, in which they pointed out the need of updating the test method regarding the method of specimen preparation, degree of saturation, air void determination, standardization, and certification. Ohio Department of Transportation has already implemented Superpave Level I volumetric based design method for developing Job Mix Formula (JMF) for HMA. In the meantime, there has been a lag behind the necessary structured laboratory experiments to ascertain the applicability of current AASHTO T 283 to Ohios Superpave mixes. This research is to conduct a structured laboratory test program using Ohio typical aggregate and asphalt binder, so that AASHTO T 283 can be either modified or improved to make it suitable for Superpave HMA.

Figure1- 1 HMA at the Time of Mixing (McCann et al., 2001)

Figure 1- 2 Displacement and Detachment of Asphalt Binder in the Presence of Moisture (McCann et al., 2001)

1.1.2 Rutting and Low Temperature Cracking Rutting or permanent deformation is a major distress affecting the pavement structures. It is the deformation of twin longitudinal depressions in the wheel paths due to a progressive accumulation of permanent deformation in one or more of the pavement layers (Anani, A. B, 1990). Permanent deformation (Christensen, W. D., and Bonaquist, R., 2002) can occur from two types of mechanical response: viscous flow and plastic deformation. Plastic deformation occurs as aggregate particles move slightly relative to one another, which is accompanied by viscous flow in the asphalt cement binding these particles together. These processes are difficult to analyze quantitatively, though simple in concept. Many Hot Mix Asphalt material properties and the strength parameters obtained by the simple performance test were used to correlate to the rutting potential by many researchers. The material properties including nominal maximum aggregate size, optimum asphalt content, performance grade of binder, voids in the mineral aggregate, voids filled with asphalt, fines to asphalt ratio, surface area of the aggregate blend, and film thickness were investigated, some trends were found. The strength parameters from the tests such as indirect tensile test, unconfined compression test, Marshall stability test, creep test were used to correlate to the rutting behavior, good relationships were obtained by some researchers, and the others were not. Most of research was based on the unmodified asphalt. Since more modified asphalt was used in the field, more research about the modified asphalt was needed. The conclusion that the modified asphalt significantly improved the rutting resistance was proved by laboratory and field test results, while, a comprehensive relationship of material properties and the strength 5

parameter was still quite elusive. This research is to identify the relationship between the characteristics of HMA using modified asphalt with its rutting behavior. Thermal or low temperature cracking of asphalt concrete pavements is a serious problem in many regions of the United States. Low temperature cracking is attributed to tensile stresses induced in asphalt concrete pavement by temperature drops to extremely low levels. If the pavement is cooled to a low-temperature, tensile stresses develop as the pavement contracts. The friction between the pavement and the base layer resist the contraction. When the tensile stress induced in the pavement equals the strength of asphalt concrete mixture at that temperature, a microcrack develops at the edge and surface of the pavement. At colder temperature or repeated temperature cycles, the crack penetrates the full depth and width of the asphalt concrete layer. Polymer modifiers have been advocated by industry to have the ability to improve performance characteristics of asphalt concrete, such as low temperature cracking. However, there also have been reports, indicating premature failures of polymer modified asphalt concrete in the field. In fact, there has a lack of comprehensive and systematic laboratory investigation of mechanical properties and performance characteristics of hot mix asphalt containing polymer modifiers. Also there is a need to investigate the HMAs fundamental mechanical property that can be linked back to HMAs low temperature cracking behavior.

1.2 Objectives of the Study The pavement distresses including moisture damage, rutting and thermal cracking are investigated in this research. 1.2.1 Moisture Damage The main objective of this study is to improve and modify the AASHTO T 283, resistance of compacted asphalt mixtures to moisture induced damage, specifically for Superpave HMA in Ohio. Specific objectives of the study are enumerated below: Conduct a structured laboratory experimental study on the various factors affecting dry tensile strength, freeze-thaw tensile strength, and tensile strength ratio (TSR). The factors to be studied include aggregate source, asphalt binder, compaction method, specimen size, aging method, degree of saturation, and freeze-thaw cycle. The laboratory experimental program is divided into two parts: one part is a complete factorial experimental program for two aggregate sources with virgin asphalt binder, the second part is a partial factorial experimental program for the other two aggregate sources with polymer modified asphalt binders. The complete factorial experiment is used to identify and evaluate the influences of the individual factors, along with all the possible interactions among the various factors. The partial factorial experiment program, on the other hand, is used to validate findings from a complete factorial experiment program. A comprehensive analysis of test data, using statistical analysis, such as ANOVA (Analysis of Variance) technique and regression analysis technique, is conducted 7

to derive quantitative grouping of the importance of various test variables. This information, in turn, will be used to enable objective modification of current AASHTO and NCHRP (SHRP) test procedures, leading to the development of specific recommendations for AASHTO T 283 test procedures for ODOT. Develop test procedures for moisture damage assessment, based on proper modification of AASHTO T283, for ODOT implementation. 1.2.2 Rutting and Low Temperature Cracking The main objective of this study is to investigate the relationship between the characteristics of HMA using modified asphalt with its rutting and low temperature cracking behavior. Specific objectives of the study are enumerated below: Explore the influence of polymer modifiers adding to the asphalt binders and mixtures to modify their properties, particularly rutting and low temperature cracking resistance. Investigate the relationship between the rutting performance of the modified asphalt concrete mixtures and their material properties and/or strength parameters. Investigate the modified asphalt concrete mixtures fundamental mechanical property that can be linked back to HMAs low temperature cracking behavior. 1.3 Organization of Dissertation The organization of dissertation is as follows: Chapter II contains the literature review of current research activities and findings about the HMAs stripping, rutting and low temperature cracking behavior. 8

Chapter III presents the complete analysis results of the complete factorial test data. Effects of various factors on the water stripping test results are given. Chapter IV concerns with analysis of the partial factorial test data and recommended water stripping test procedures for Superpave HMA specimens. Chapter V deals with rutting behavior of HMA containing polymer modifiers. A new prediction model to estimate the pavement rut depth using HMAs physical properties and mechanical properties is given. Chapter VI contains low temperature cracking behavior of HMA containing polymer modifiers. Several trends are found and preliminary conclusions are given. Chapter VII presents summaries and conclusions of the research work.

Recommendations for the future research are provided at the end of the chapter.

CHAPTER II LITERATURE REVIEW

2.1 Literature Review of Moisture Damage In the late 1970s and early 1980s, a significant number of pavements in the United States began to experience distress associated with moisture sensitivity of hot mix asphalt (HMA) materials. Rutting, raveling, and cracking due to moisture induced damages were observed on many pavements. The causes of this sudden increase in pavement distress because of water sensitivity have not been conclusively identified. Regardless of the cause of this moisture related premature distress, methods are needed to identify HMA behavior in the presence of moisture. Test methods and pavement performance prediction tools need to be developed that couple the effects of moisture on the properties of HMA mixtures with performance prediction to estimate the behavior of the mixture in resisting rutting, fatigue, and thermal cracking when it is subjected to moisture under different traffic levels in various climates. Most state highway agencies have implemented the Superpave (Superior Performing Asphalt Pavement) mix design system developed through the five year research effort of Strategic Highway Research Program (SHRP). This mix design method allows agencies or contractors to design asphalt mixes that meet certain performance requirements as

10

dictated by traffic, environment, and location of the pavement system (Cominsky et al., 1994). Superpave mixes have typically been coarse-graded (gradation passing below the maximum density line) in order to provide a greater volume of stone in the mix to aid in reducing the potential for rutting, fatigue and stripping, and to provide stronger HMA pavement layer than that designed in the conventional methods using Marshall mix design (Hall et al., 2000). The Superpave level I mix design (volumetric based design) is based on gyratory compactor, as shown in Figure 2-1, which typically produces 6 inch diameter specimens. 2.1.1 AASHTO T-283 Test Method Several laboratory tests have been developed to assess the moisture susceptibility of HMA. Laboratory testing does not entirely simulate field conditions; however, it can provide useful information. These tests developed for the evaluation of moisture damage either assess the stripping of asphalt from the aggregate surface or the loss in the strength of the compacted HMA specimens. Currently, AASHTO T 283 Specification is the most widely used method to evaluate the HMA stripping potential. The T 283 procedure consists of preparing 6 HMA samples using Marshall impact compaction method. A picture of the Marshall Impact Compactor is shown in Figure 2-2. The air voids of the prepared samples are between 6 and 8 percent. The required high percentage of air voids helps accelerate moisture damage to the HMA specimens. The samples are divided into two groups: the first group is the control group, or unconditioned, while the second group, or conditioned, is vacuum-saturated (a picture of the device for sample saturation is shown in Figure 2-3) 55-80 percent 11

(AASHTO T283-99) or 70-80 percent (AASHTO T283-03) with water and then placed in a freezer at 0 F for 16 to 18 hours.

The conditioned specimens are then placed in a water bath (A picture of the device for bathing is shown in Figure 2-4) at 140 F for 24 hours. After the freeze/thaw

conditioning is done, the indirect tensile strength (St) measured by MTS machine (A picture of MTS machine is shown in Figure 2-5) or a simpler machine is determined for all samples with a loading rate of 2 in/min. The tensile strength of conditioned sample St(Conditioned) is compared to the tensile strength of unconditioned sample St(Control) to determine Tensile Strength Ratio (TSR) as follows:

TSR=

S t(Conditioned) S t(Control) (Eq. 1)

A visual (subjective) estimation of the magnitude of stripping of the conditioned sample completes the test procedure. In summary, most agencies use a minimum value of TSR = 80% in the moisture sensitivity test for Superpave HMA mixtures.

12

Figure 2-1 The Superpave Gyratory Compactor

13

Figure 2-2 The Marshall Impact Compactor

14

Figure 2-3 Saturation Chamber

Figure 2-4 Water Bath Container 15

Figure 2-5 MTS Machine

2.1.2 Efforts to Develop Alternative Test Methods

Buchanan et al. (2004) summarized some of the disadvantages of the AASHTO T 283:
Performance of T 283 test is time consuming. The wide range of saturation level (55% - 80%) may result in substantial TSR

variability.
Uncertainty of test results on the Superpave specimens due to the difference in the

Superpave specimen size of 6 inches diameter, as contrast to the 4 inch diameter Marshall samples in the T 283 Specifications.
Visual (subjective) examination is required to estimate the magnitude of the

conditioned sample stripping. 16

 The conditioning procedure in T 283 does not simulate repeated generation of pore

pressure under loads, which is believed to be a major cause of stripping in pavements. As a result, there have been numerous studies focusing on investigating new testing methods or procedures that can characterize the water sensitivity in Superpave HMA, as a substitute to the AASHTO T 283. Cross et al. (2000) used the loaded wheel tester of Asphalt Pavement Analyzer (APA) to detect moisture susceptible mixtures. Eight different mixes from seven project sites were evaluated with APA. Samples were tested using four different preconditioning procedures: dry, soaked, saturated, and saturated with a freeze cycle. The results indicated that the APA could be utilized to evaluate the moisture susceptibility of asphalt mixes. Additionally, the results indicated that harsh preconditioning of saturation or saturation with a freeze cycle did not result in the increased wet rut depth. Pan et al. (1998 and 1999) used Purdue Wheel tracking device (PURWheel) to predict the conditions that promote water stripping of HMA specimens. Two major variables were selected for the research, aggregate type and anti-stripping additive. The PURWheel proved to be an effective tool for evaluating HMA mixture stripping potential in a hot/wet environment. The PURWheel also is shown capable of evaluating HMA mixture hot/wet rutting potential. Pan et al. (1998 and 1999) also pointed out that results from AASHTO T283 indicated that moisture conditioning has a significant effect on the stripping potential of the seven mixes tested. The tensile strength of the mixture was reduced after subjected to the environmental conditioning procedures in AASHTO T283 tests. Laboratory wheel tracking test results for all seven types of mixtures indicated that 17

temperature and moisture conditions were significant. It is obvious that damage occurs much faster with wet conditions. Both factors are important in identifying asphalt mixture stripping/rutting potential. Aggregate type has a significant effect on the wheel track test results. Limestone generally provided better performance than other types of aggregate. McCann et al. (2001) used the ultrasonic moisture accelerated conditioning process to quantify the moisture sensitivity of HMA pavement. A total of 13 HMA mixtures were subjected to ultrasonic moisture accelerated conditioning. The mixtures represented typical Hveem or Superpave mixtures with a 25 mm maximum size aggregate used by the Nevada Department of Transportation. For the laboratory assessment of moisture sensitivity within the mixes, variables included grade of asphalt binder, percent asphalt binder, aggregate type, and mixes with and without lime. Test results from ultrasonic conditioning were then compared to tensile strength ratios derived from the conventional testing procedures. The main conclusion of the research was that the research hypothesis the loss of material is proportional to the length of time a HMA sample is subjected to ultrasonic conditioning is found to be true. For the HMA mixes subjected to ultrasonic conditioning and analyzed using linear regression, differences as to aggregate source, type of binder, lime used as an aggregate additive, and the percent asphalt binder within the mix could be detected. The determination for the potential of stripping by ultrasonic moisture accelerated conditioning is analogous to results established by tensile strength testing after 18 cycles of freeze-thaw conditioning. In Mississippi, Buchanan et al. (2004) evaluated the use of Moisture Induced Stress Tester (MIST) for moisture sensitivity tests and compared its results with that of AASHTO T 283. Basically the MIST simulates HMA pavement stripping mechanism by 18

using compressed air to force water out of the HMA and then by depressurizing (creating vacuum) to pull back water into the specimen. By repeating the pressure/depressurization cycles, stripping in a form of emulsification of the asphalt binder and stripping of the fine aggregate from HMA will ensue, resulting in an increase in the turbidity of water. By measuring the scattered/transmitted turbidity ratio, it was shown that the MIST was capable of detecting adhesion failure (stripping). In Florida, Birgisson, et al. (2004) found that moisture damage has an impact on the fracture resistance of mixtures that is accurately captured by the fundamental parameters of the HAM fracture mechanics model. This means that HAM fracture mechanics can be used to quantify the effects of moisture damage on mixtures. Based on the detailed forensic investigations of 36 field pavement sections of known cracking performance in Florida, a HAM fracture mechanics-based performance specification criterion, termed the Energy Ratio (ER), was used to quantify the effects of moisture damage on the fracture resistance of mixtures. Based on Birgisson et al. results the ER was recommended to form the basis of a promising specification criterion for evaluating the effects of moisture damage in mixtures as well as the overall resistance to fracture. Khosla et al. (2000) investigated an alternative test that evaluates a mixtures fundamental material properties instead of measuring indirect tensile strength ratio. A relative simple test is proposed that measures the cohesion and friction angle for asphalt mixture. The Superpave shear tester was incorporated as a tool in moisture sensitivity evaluation. Based on the tests results, they found that the proposed test apparatus provides a simple method for determining the cohesion and friction angle of a mixture and may be a new way to evaluate the moisture sensitivity of the mixture. 19

2.1.3 Previous Experience in Modifying the AASHTO T 283

Some agencies and researchers have investigated AASHTO T 283 and modified the test procedures. 2.1.3.1 NCHRP 444 report Compatibility of a Test for Moisture-Induced Damage with Superpave Volumetric Mix Design This NCHRP 444 project was aimed at evaluating the AASHTO T 283 specification and recommending changes to make it compatible with the Superpave system. Comparisons of the test procedures of ASTM D4867, AASHTO T 283 and Superpave are presented in Table 2-1 (NCHRP 444, 2000; AASHTO Specification Book, 2001; and ASTM Standards, 2001). The differences between T 283 and the Strategic Highway Research Program (SHRP) recommended AASHTO T 283 for Superpave mixtures include the time and temperature of aging and the size of the HMA sample (diameter and height). The SHRP research, however, was deficient due to insufficient testing to establish better understanding of the TSR in relation to sample preparation methods, such as sample conditioning, method of compaction, and size of sample. In general, the TSR ratios of Superpave 6 inch diameter specimens were larger than the Superpave 4 inch specimens. These differences in the TSR are due to the generally higher St(Control) of dry specimens and lower conditioned St(Conditioned) obtained on the Superpave 4 inch specimens as compared with the Superpave 6 inch specimens. On the other hand, the TSR for the 4 inch Marshall compacted specimens is similar to the TSR obtained for the Superpave 6 inch specimens. One of the report final conclusions stated that there is little difference in variability of test results among methods of compaction. A modified AASHTO T 283 method was introduced as a result of this NCHRP study. This method 20

allows the use of both 4 and 6 inch specimens depending on the aggregate size used in the HMA mixture. The saturation level range was expanded to 50-80 %, instead of 55-80 % in the standard AASHTO T 283. The aging of loose mixture and freeze/thaw cycle were recommended as well. Nevertheless, the major recommendation in this study for the state agencies was that they should carry out their own experimental HMA moisture damage testing to determine the comparative behavior of their aggregates and binder sources before switching to the Superpave 6 inch samples. 2.1.3.2 Specimen size NCHRP 444 recommends using the Gyratory specimen of both 4 and 6 inch specimens. UTDOT and NCDOT use the Gyratory specimen of the 6 inch-diameter specimens. FDOT uses the 4 inch-diameter specimens, while NYDOT chooses the size based on the nominal maximum aggregate size. 2.1.3.3 Air void and saturation level NCHRP 444 recommends that saturation level range should be expanded to 50-80%, instead of 55-80 % in the standard AASHTO T 283. Khosla et al. (2000) concluded that a mix at 6% air voids and 55% saturation level may pass TSR requirement, but could fail at a higher level of air voids and degree of saturation, even though both of them are within the AASHTO T-283 specifications. They recommended tightening the standards range for air voids and degree of saturation. Choubane et al. (2000) investigated the effects of different degrees of saturation on moisture damage. When using AASHTO T283 for moisture susceptibility evaluation of Superpave mixes, it is recommend that the test samples be saturated to more than 90%. It is also suggested that the conditioning phase include the optional freeze-thaw cycle. An 21

appropriate passing TSR limit should be set to no less than 80%. The air void content of test samples should be reduced to 6.5%-7.5%. Castro-Fernandez et al. (2004) found that after 24 hours of water conditioning process at 140 F, the loss in aggregate internal friction and the loss of aggregate-asphalt binder adhesion (retained compressive strength) are more prevalent than the loss of asphalt binder cohesion. 2.1.3.4 Loose mix aging and compacted mix aging Most of the modified AASHTO T 283 required loose mix aging and compacted mix aging. The loose mix aging was 2 hours at compaction temperature. The compacted mix aging was 24-96 hours at room temperature, except for UTDOT who used 16 hours at 140 F. 2.1.3.5 Freeze-thaw cycle Most of the modified AASHTO T 283 required the freeze/thaw cycle. 2.1.3.6 Minimum TSR FDOT early experience suggested a minimum TSR value of 85 % for Superpave mixtures (Musselman et al., 1998). Later on, Choubane et al. (2000) recommended the use of minimum TSR value of 80 % for water damage evaluation in FDOT Superpave HMA. Also, Khosla et al. (2000), McGennis et al. (1996), and Pan et al. (1999) suggested a minimum value of TSR = 80%. In summary, most agencies use a minimum value of TSR = 80% in the moisture sensitivity test for Superpave HMA mixtures. As it was discussed previously, many attempts by state agencies had been made to modify the AASHTO T 283 standards. The collected information about these efforts from different state agencies is summarized in Table 2-2. 22

2.1.4 Summary

The current Superpave specification uses the AASHTO T 283 moisture susceptibility test for determining moisture sensitivity of HMA specimens. Most state agencies use AASHTO T 283 test, although they still have questions about the accuracy of the test. There is a need to develop a verified test procedure that can be adopted for determining moisture damage resistance of Superpave specimens. The uncertainties surrounding the AASHTO T283 and NCHRP 444 procedures, particularly regarding their impacts on Ohios aggregate sources and ODOT Superpave procedures, provided impetus for carrying out this research. Despite the fact that there are many different test methods for moisture sensitivity of HMA specimens, Hicks et al. (2004) pointed out the need of updating the moisture damage susceptibility test method, particularly regarding the method of specimen preparation, degree of saturation, air void determination, and test procedure standardization. Therefore, a need exists for gaining better understanding of the important factors controlling the accuracy and validity of the stripping test. In this study, a structured laboratory test program is conducted to study the effects of the dominant test variables as well as potential interactions between these test variables on the HMA specimens susceptibility to moisture damage. The factors studied include aggregate source, compaction method, specimen size, aging method, saturation level, and freezethaw cycle. The effects of these factors on the dry tensile strength, the conditioned tensile strength, and the tensile strength ratio are fully investigated and documented in this research.

23

Table 2-1 Comparison of Standard Water Sensitivity Tests

AASHTO T 283-99 (AASHTO T 283-03) 2.5 x 4 Loose mix aging None Cool @ room temp. (2 hrs). Cure @ 140o F -16 hrs. Compacted HMA curing Compaction Temperature Air voids of compacted specimen Saturation 0-24 hrs @ room temperature before staring test Depends 24 hrs @ room temp. before starting test 2750 F (2 hrs in oven) Same as AASHTO T 283 Equiviscous (0.28 Pa.s) 6-8% 275o F 4 hrs Superpave (Recommended by SHRP) 3.75 x 6

Test Parameter

ASTM D4867

Specimen size

2.5 x 4

6-8%

6-8% (6.5-7.5%) 55-80% (70-80%) 20 in. Hg for 5 min. Calculations different from ASTM D4867

55-80% 20 in. Hg for 5 min. Calculations different from AASHTO T 283

Same as in AASHTO T 283

Swell determination Freeze

Yes 0 5 F for min. 15 hrs (optional) 140 + 0.2F for 24 hrs Marshall Indirect Tensile @ room temp

No 0 5 F for min. 16 hrs (optional) 140 + 0.2 F for 24 hrs Marshall Indirect Tensile @ room temp

No Same as in AASHTO T 283 140 + 0.2 F for 24 hrs SGC Same as in AASHTO T 283

Water soak Compaction method

Strength property

24

Table 2-2 Summary of State Agencies Experience in Modifying the AASHTO T 283 in different U.S. Departments of Transportation

Specimen Size (in.)

Compaction method

Air voids % Same

Saturation %

Loose Mixture Aging

Compacted Mixture Aging 0-24 hrs @ room Temp.

Freeze/ Thaw

Min. TSR%

Reference

4 or 6

Marshall or SGC

as in T283 Same

Same as in T - 283

Same as in T - 283

Compulsory

80

NCHRP 444 (2000)

2 hrs +- 5 70-80 min. @ compaction temp. Compulsory 290+-5 F for 2hrs 24-96 hrs in room temp. (different approaches for 4" and 6") Visual estimation NCDOT (1999) 16 hrs @ 140 F Compulsory N/A UTDOT (2003)

SGC

as in T283 4"

Marshall for 4 or 6 4" or SGC for 6"

(same as in T283). 6"(6.57.5) Same 65-80

SGC

as in T283

70-80

Same as in T - 283

N/A

Compulsory

N/A

FDOT (2002)

12 Specimens (4")

CA method similar to Marshall

6.57.5 Same

70-80

Same as in T - 283

24-96 hrs @ room temp. Compulsory N/A

CADOT (2003)

Marshall

as in T283

Same as in T - 283

Short term aging for 2 hrs

Same as in T - 283

Same as in T - 283

80

INDOT (1999)

Based on nominal max. agg. size SGC

Same as in T283 Same as in T - 283 Same as in T - 283

Up to 96 hrs @room temp.

Same as in T - 283

80

NYDOT(2002)

25

2.2 Literature Review of Rutting One of the most serious types of structural distresses of asphalt concrete is the permanent deformation, which could be referred to as rutting. Permanent deformation or rutting has been and continues to be a problem in the performance of hot mix asphalt pavements. In recent years, the potential for rutting on the nations highway has increased due to higher inflation pressures. Therefore, the accurate understanding and prediction of the mechanisms involved in the rutting phenomenon is an essential issue for an efficient management of pavement system. Many mixture characteristics are believed to significantly affect pavement rutting, though a comprehensive relationship of such characteristics to rutting has been proven to be quite elusive (Stacy G. Williams, 2002). Rutting predictions by physical properties of HMA and mechanics properties of HMA are desirable. They can be used to screen and differentiate rutting properties. Experience shows that stiff binders with large aggregates typically are more resistance to rutting than mixes containing finer aggregates and higher binder content (Kandhal, P. S., 2001). Polymer-modified binders can be used to improve rutting performance. Coarse, angular aggregates are thought to have more rutting resistance than fine angular ones. In general, rutting resistance can be increased through the use of proper aggregate sizes and shapes, as well as appropriate binder selections. A variety of laboratory test methods have been developed in order to gain more accurate prediction of the permanent deformation characteristics of pavements. Some of the tests are used for many years, while others are still in the developmental stage. From the literature review, the strength-based laboratory test methods such as indirect tensile 26

test, creep test, unconfined compression test, Hamburg wheel test, and asphalt pavement analyzer have been used to evaluate rutting of asphalt concrete mixes. Invariably, these tests have attempted to simulate vertical and lateral stress and compatibility conditions of an asphalt pavement under temperature-controlled conditions.
2.2.1 Previous Findings Based on the Unmodified Asphalt:

1. Rutting Prediction Based on the Physical Properties of HMA

Stacy (2002) explored the relationships of HMA mixture properties to rutting susceptibility as measured by utilization of wheel-tracking devices. The conclusion was that many factors such as: VMA, PG binder grade, and binder content played a role in the rutting characteristics of HMA samples, regression procedures were unable to determined valid mathematical relationships; however, several trends were found: as VMA, binder content increased, so did rut depth; as the PG grade increased, rut depth decreased.

Thiessen (2000) investigated the relationship between material properties and strength parameters measured in the laboratory to observed rutting behavior in the field. Physical mixture properties can not, by themselves, explain the observed rutting performance of the pavements tested. Regression analysis of the both the as-constructed and recovered core properties show poor correlation to the rate of rutting. Of the properties examined the recovered binder penetration and percent aggregate passing the No. 4 sieve showed the strongest correlation to rutting rate. Other properties such as air voids and asphalt content showed little

27

correlation to rutting rate, although experiences would suggest a relationship does exist.

2. Rutting Prediction Based on the Mechanics Properties of HMA

Zhu (2001) investigated feasibility of replacing wheel tracking test by a repeatedload creep test for rutting potential evaluation. Favorable conclusions concerning the use of the repeated-load creep test for rutting potential were drawn based on the findings of the experimental test results. The correlation between the two types of tests was found to be good for all three asphalt mixtures investigated.

Thiessen (2000) investigated the rutting performance of asphalt pavements in Manitoba. Samples taken from the filed were tested using modified indirect tensile test, results of analysis clearly showed that measured deformation and Poissons ratio obtained from the indirect tensile test are closely related to rutting in the field.

Christensen (2002) used a simple method for evaluating the rut resistance by indirect tensile test and unconfined compressive test. His research shown surprisingly strong correlations between these data and the rut resistance of asphalt concrete mixtures as observed in the field.

Anderson (2003) found that IDT strength not only provided an excellent measure of mixture cohesion and also correlated very well both to laboratory indicators of rutting resistance and measured rutting rates in actual pavement.

3. Rutting prediction based on both of physical and mechanics properties of HMA

28

Zaniewski (2004) used IDT strength and voids in mineral aggregate to correlate rutting potential measured with Asphalt Pavement Analyzer (APA) and a strong correlation was found.

2.2.2 Previous Findings Based on the Modified Asphalt:

Bilal (2004) investigated the influence of modifiers to the rut depth measured by Asphalt Pavement Analyzer (APA). He found that asphalt concrete mixtures contained SBS modifiers and SBR Butanol significantly improved the rutting resistance compared to unmodified asphalt concrete mixture.

Pedro (2006) found that SBS modified binders outrank stone matrix asphalt gradation in plastic deformation performance. Aggregate source, gradation and asphalt content were found of very little effect, in comparison to the addition of SBS. When unmodified binders were used, nevertheless, aggregate source, gradation and type of asphalt binder became more important.

2.3 Literature Review of Low Temperature Cracking Thermal cracking of asphalt pavement is a common phenomenon in cold regions and has been a recognized problem for some time in Canadian province and the northern United States. This type of distress is manifested as a series of transverse cracks that extend across the pavement surface in response to cold ambient temperature. The mechanism associated with low temperature cracking is simply that tensile stresses develop in an asphalt concrete pavement as the temperature drops to an 29

extremely low value. When the tensile strength is equal to the tensile strength of the mixture at a given temperature, a micro crack develops at the surface of the pavement. The crack then propagates through the depth of the layer when subjected to additional thermal cycles. A number of test methods have been used to study the phenomena of low temperature cracking in asphalt concrete mixes.
Indirect Diametral Tesion Test Direct TensionConstant Rate of Extension Tensile Creep Test Flexural Bending Test Thermal Stress Retained Specimen Test Three-Point Bend Specimen Test C*-Line Integral Test Coefficient of Thermal Expansion and Contraction Test Determination of Thermal Conductivity and Specific Heat Capacity

Based on the evaluation of the test methods by Vinson et al. (1990), TSRST was judged to have the greatest potential to evaluate the low temperature cracking susceptibility of an asphalt concrete mixture. The test has been successfully used by several investigators to characterize the response of asphalt concrete mixtures at low temperatures (Monismith et al.1965, Fabb 1974, Carpenter 1983, Arand 1987, Sugawara et al. 1982, Janoo 1989). Polymer-modified binders can improve low temperature cracking resistance, however, the research is limited. There has a lack of comprehensive and systematic laboratory 30

investigation of mechanical properties and performance characteristics of hot mix asphalt containing polymer modifiers. Also there is a need to investigate the HMAs fundamental mechanical property that can be linked back to HMAs low temperature cracking behavior. In SHRP-A-400, an experiment design that considered a range of mixture and test condition variables was developed to evaluate the suitability of TSRST for characterizing low temperature cracking resistance of asphalt concrete mixtures. Four asphalts and two aggregates were selected for the experiment. TSRST results provide an excellent indication of the low temperature cracking resistance of asphalt concrete mixtures. All the variables considered in the experiment designs are significant factors in the TSRST results. The penetration of asphalt cement at 25C combined with the viscosity at 60C or the ring and ball (R&B) softening point provide an excellent possibility for predicting the fracture temperature of asphalt concrete mixtures. The penetration index and penetration viscosity number did not provide a good definitive relationship with fracture temperature. Peter E. Sebaaly et al. (2002) summarizes Nevada's experience with the Strategic Highway Research Program low temperature tests and specifications. An investigation of the Superpave Performance Graded Binder tests has determined that the bending beam rheometer and the direct tension test correlate very well and it may not be necessary to run both tests, as they are set up in the current Superpave specifications. The TSRST appears to provide the greatest value for evaluating low temperature properties of HMA mixtures. Findings from the research indicate that there are some significant correlations between the low temperature properties of asphalt binders and HMA mixtures if the mixtures are aged appropriately. On the other hand, the research showed that, when using 31

polymer-modified asphalt binders, the low temperature grade of the asphalt binder may be conservative enough that testing of the HMA mix may not be necessary. Timothy R. Clyne et al. investigated of the low temperature fracture properties of three MnROAD asphalt mixtures. In this research effort, field cores were taken from cells 33, 34 and 35 at the Minnesota Department of Transportation (MnROAD) facility to determine the fracture properties of the field mixtures, to compare them with the laboratory-prepared mixtures analyzed in a previous study, and to evaluate the effect of aging at different depths in the asphalt layer. In addition, the properties of the recovered binders from the field cores as well as the properties of the original binders aged in laboratory conditions were investigated. The test results and the analyses performed indicate that the fracture tests performed on asphalt binders and asphalt mixtures have the potential to predict the field performance of asphalt pavements with respect to thermal cracking. In NCHRP 465, the research team conducted a comprehensive laboratory testing program to statistically correlate the actual performance of HMA materials from the MnRoad, Wes-Track, and FHWA Accelerated Loading Facility (ALF) experiments with the measured responses of specimens prepared from original materials for 33 promising test methodtest parameter combinations. For low temperature cracking, three laboratory tests had parameters that resulted in a fair correlation to the measured amount of cracking. These test methods and the mixture response parameters were the dynamic modulus measured from the triaxial compression test at lower test temperatures, the compliance measured at 1,000 s from the indirect tensile creep test, and the tensile strain at failure adjusted by the modulus of the mixture. The tensile strain at failure is determined from 32

the indirect tensile strength test, and the modulus from the triaxial dynamic modulus test. The team recommends the creep compliance measured by the indirect tensile creep test at long loading times and low temperatures and this recommendation is based solely on work carried out for SHRP and C-SHRP and recently con- firmed in NCHRP Project 137A.

33

CHAPTER III FACTOTS AFFECTING WATER STRIPPING TEST RESULTS

3.1 Introduction The objectives of this study are to evaluate AASHTO T283, Resistance of Compacted Bituminous Mixture to Moisture Induced Damage, and to recommend changes to make it compatible with the Superpave mix design procedures. The central issue of the study is to determine the influences of the compaction method and size of sample on the results of the AASHTO T283 method of test. Comparisons of indirect tensile strength test results have been made among the 150-mm-diameter Superpave gyratory HMA samples, 100mm-diameter Superpave gyratory HMA samples, and 100-mm-diameter Marshall HMA samples. The influences of compaction method and sample size are analyzed in terms of dry tensile strength, freeze-thaw conditioned tensile strength, and tensile strength ratio. In the data analysis, pair-wise comparisons are used to compare the impact of each individual factor while maintaining all other factors at a constant level. This type of analysis is an excellent tool to identify the effects of each individual factor in terms of direction (increase or decrease) and magnitude. For example, using the pair-wise analysis, an engineer can assess whether the dry tensile strength of G150 samples is equal to, lower than, or higher than the dry tensile strength of M100 samples. Criterion of 95%

34

confidence is used in data analysis. The significance value less than 0.05 is considered statistically different. (ANOVA) test and regression analysis technique are used in analyzing test data as well. The analysis of variance (ANOVA) technique is used to conduct the statistical analysis of the overall data generated from the complete factorial experiments. The contributions of the main factors to the water stripping test results and their interactions are identified by ANOVA analysis techniques. The magnitude of the Fstatistics is used to rank the relative importance of the influencing factors and their interactions. Regression analysis technique is used to evaluate the tensile strength ratio for different compaction methods. 3.2 Experimental Program A structured laboratory test program is designed and carried out in this research. The factors selected for the test program are based on the literature review.
3.2.1 Introduction of Research Approach

The complete factorial experiment is use to identify and evaluate the contributions of the individual factors, along with all the possible interactions among the various factors. The partial factorial experiment is used to validate findings from a complete factorial experiment test. Four aggregate sources are used in the laboratory experimental study: one limestone, Traprock, and two gravels. Two types of asphalt binders are used: one is virgin binder (PG 64-22), and the other is pre-blended polymer modified asphalt binder (PG 70-22). The test variables to be included in the test program encompass the following: loose mixture aging (none versus AASHTO T283, 16 hrs @ 140 F, as well as Superpave 35

specifications, 2 hrs @ 275 F or 4 hrs @ 275 F), methods of specimen compaction (Marshall versus Superpave gyratory), specimen sizes (4 inch for Marshall, 4 inch and 6 inch for Superpave gyratory), aging condition of compacted HMA specimens (0-24 hrs versus 72-96 hrs @ room temp.), degree of saturation (three levels: 55, 75, and 90%), and freezing thawing condition (none versus standard one freeze/thaw cycle @ 16 hrs @ 0 F). Previously in Chapter II, Table 2-1 provides details of current ASTM, AASHTO, and Superpave test conditions; these are the basis of test variables identified above. The test program is divided into two parts: one part is a complete factorial experimental program for two aggregate sources with virgin asphalt binder, the second part is a partial factorial experimental program for the remaining two aggregate sources with modified asphalt binder. Figure 3-1 depicts the independent test variables in a complete factorial laboratory test program. The number of tests to be conducted can be roughly estimated as 3 (replicates) 2 (aggregate source) 3 (loose mixture aging) 3 (specimen size) 4 (degree of saturation) 2 (w and w/o freeze thaw conditioning) = 864 specimens. The partial factorial experimental program is based on the conclusions from the analysis of the complete factorial experimental program. The complete factorial experiment is used to identify and evaluate the influences of the individual factors, along with all the possible interactions among the various factors. The partial factorial experiment is used to validate findings from a complete factorial experiment test.
3.2.2 Materials Tested

Two different aggregates from two districts in Ohio are used in the complete factorial experimental program. One source is the limestone from the Honey Creek Stone Co. in 36

Petersburg, Ohio. The other source is the gravel from Martin Marietta Co. in Ohio. Madhar conducted the experimental tests for this source. I really appreciated his help. Another two different aggregates from two districts in Ohio are used in the partial factorial experimental tests. One source is the Traprock from the Ontario Trap Rock Co. in London, Ontario, Canada. The other source is the gravel from Stocker Sand and Gravel Co. in Gnadenhutten, Ohio. All the aggregate gradation used is according to ODOT requirements for heavy traffic (Type-1H) as shown in Figure 3-2. The virgin asphalt binder used in the complete factorial experimental program is a performance grade PG 64-22 (from Tri-State Co.). The Superpave asphalt binder specification (AASHTO MP1-93) for the PG 64-22 is shown in Table 3-1. The modified asphalt binder used in partial factorial experimental test is a performance grade PG 70-22 provided by Marathon Petroleum Company. The AASHTO T 283 procedure consists of preparing 6 HMA samples for one set. Each set of specimens is divided into subsets. One subset named unconditioned is tested in dry condition for indirect tensile strength. The other subset named conditioned is subjected to vacuum saturation and a freeze cycle followed by a warm-water soaking cycle, before being tested for indirect tensile strength. All specimens are tested for indirect-tensile strength at 77 F using a loading rate of 2 in/minute, and the Tensile Strength Ratio (TSR) is determined. A minimum TSR of 0.8 is usually specified. The test results including values of dry tensile strength, freeze-thaw tensile strength, and tensile strength ratio for complete factorial experimental program test are listed in Appendix A. The variability of test results is statistically examined via. mean, standard deviation, and coefficient of variation (COV). The mix designs according to Marshall and 37

Superpave procedures for all four aggregate sources with the associated asphalt binder specifications are summarized in the Appendix B.

Figure 3-1 Complete Factorial Experimental Test Program

38

Aggregate Gradation Curves of the Upper and Lower Limits of the ODOT Type-1H and the Job Mix Formula Used in this Research
120

100

80

Percen Passing

Lower Limit

60

Upper Limit JMF

40

20

0 No. 200 No. 100 No. 50 No. 30 No. 16 No. 8 No. 4 3/8" 1/2" 3/4"

Sieve Size

Figure 3-2 ODOT Requirements for Heavy Traffic (Type-1H)

Table 3-1 Superpave Asphalt Binder Specification for the PG 64-22 Average 7-Day Maximum Pavement Design Temperature (C). Minimum Pavement Design Temperature (C) Flash Point Temp. Minimum. (C) Temperature at Maximum Viscosity of 3000 cP. (C) <64 >-22 230 135

39

3.3 Data Analysis and Discussion Statistical methods including pair-wise comparisons, ANOVA test and regression test are used to analyze the data.
3.3.1 Dry Tensile Strength

Dry tensile strength is the indirect tensile strength of unconditioned specimen obtained by indirect tensile strength tests. 3.3.1.1 Comparison of three compaction methods Tables 3-2, 3-3, and 3-4 show statistical comparisons of samples prepared with the Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 150-mm specimen, respectively.
M100 Samples versus G100 Samples

Table 3-2 indicates that the dry tensile strengths of HMA samples compacted by the M100 compactor is statistically the same as those samples compacted by the G100 compactor. This is true for both aggregate sources.
M100 Samples versus G150 Samples

Table 3-3 indicates that the dry tensile strengths of samples compacted by the M100 compactor are statistically larger than those samples compacted by the G150 compactor for limestone. The dry tensile strength is statistically the same for gravel.
G100 Samples versus G150 Samples

Table 3-4 indicates that the dry tensile strengths of samples compacted by the G100 compactor are statistically larger than those samples compacted by the G150 compactor for limestone. The dry tensile strength is statistically the same for gravel. 40

Table 3-2 Statistical Comparison of 100-mm Diameter Marshall Compacted Samples and 100-mm Diameter Superpave Gyratory Compacted Samples Sample conditioning Source Honey Creek Dry Limestone Martin Marietta Gravel Larger* Same*

Smaller*

Table 3-3 Statistical Comparison of 100-mm Diameter Marshall Compacted Samples and 150-mm Diameter Superpave Gyratory Compacted Samples Sample conditioning Source Honey Creek Dry Limestone Martin Marietta Gravel Larger*

Same*

Smaller*

Table 3-4 Statistical Comparison of 100-mm and 150-mm Diameter Superpave Gyratory Compacted Samples Sample conditioning Source Honey Creek Dry Limestone Martin Marietta Gravel Larger*

Same*

Smaller*

41

3.3.1.2 Dry tensile strength subjected to loose mix aging and compacted mix aging Tables 3-5 and 3-6 show the influence of loose mix aging on the dry tensile strength for the Honey Creek Limestone and Martin Marietta Gravel, respectively. Tables 3-7 and 3-8 show the influence of compacted mix aging on the dry tensile strength for the Honey Creek Limestone and Martin Marietta Gravel, respectively. For Honey Creek Limestone, Table 3-5 indicates that loose mix aging increases the dry tensile strength in all six cases considered. Table 3-7 shows that compacted mix aging increases the dry tensile strength in 2 of 9 possible conditions. In 7 of 9 comparisons, the dry tensile strengths are the same. For Martin Marietta Gravel, Table 3-6 indicates that loose mix aging increases the dry tensile strength in all six cases considered. Table 3-8 shows that compacted mix aging increases the dry tensile strength in 2 of 9 test conditions. In 5 of 9 comparisons, the dry tensile strengths are the same. In 2 of 9 conditions, compacted mix aging tends to decrease the dry tensile strength.

Table 3-5 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Loose Mix Aging (Honey Creek Limestone) Compaction Method Marshall 100 Gyratory 100 Gyratory 150 Total Compacted Mix Aging Hrs 24 72 24 72 24 72 All Increase*

Same*

Decrease*

42

Table 3-6 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Loose Mix Aging (Martin Marietta Gravel) Compaction Method Marshall 100 Gyratory 100 Gyratory 150 Total Compacted Mix Aging Hrs 24 72 24 72 24 72 All Increase*

Same*

Decrease*

Table 3-7 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Honey Creek Limestone) Compaction Method Loose Mix Aging Hrs 0 4 16 0 4 16 0 4 16 All

Increase*

Same*

Decrease*

Marshall 100

Gyratory 100

Gyratory 150 Total

43

Table 3-8 Statistical Comparison of Dry Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Martin Marietta Gravel) Compaction Method Marshall 100 Loose Mix Aging Hrs 0 4 16 0 4 16 0 4 16 All

Increase*

Same*

Decrease*

Gyratory 100

Gyratory 150 Total

3.3.1.3 ANOVA analysis of dry tensile strength

ANOVA analysis of dry tensile strength for Honey Creek Limestone

Tables 3-9, 3-10, and 3-11 show the results of ANOVA analysis of the dry tensile strength of Marshall 100-mm specimen, ANOVA analysis of the dry tensile strength of Gyratory 100-mm specimen, and ANOVA analysis of the dry tensile strength of Gyratory 150-mm specimen, respectively. Table 3-12 provides the summary and comparison results of ANOVA analysis results for different compaction methods for Honey Creek Limestone. It can be seen from Tables 3-9, 3-10, 3-11 and 3-12 that the loose mix aging is the most important factor to influence the dry tensile strength. Furthermore, compacted mix 44

aging is the second important factor. The above observations are true for the Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 150-mm specimen for Honey Creek Limestone.

Table 3-9 ANOVA Analysis of the Dry Tensile Strength (Marshall 100-mm specimen) Rank 1 2 3 Source LMA CMA LMA*CMA F Value 564.211 60.677 11.116 Pr>F 0.000 0.000 0.000 Significant Y Y Y

Table 3-10 ANOVA Analysis of the Dry Tensile Strength (Gyratory 100-mm specimen) Rank 1 2 3 Source LMA CMA LMA*CMA F Value 450.665 5.223 .479 Pr>F 0.000 0.041 0.631 Significant Y Y N

Table 3-11 ANOVA Analysis of the Dry Tensile Strength (Gyratory 150-mm specimen) Rank 1 2 3 Source LMA CMA LMA*CMA F Value 178.604 21.367 7.036 Pr>F 0.000 0.000 0.001 Significant Y Y Y

45

Table 3-12 Summarizing Comparison ANOVA Analysis for the Dry Tensile Strength Compaction Rank Method M100 1 G100 G150 M100 2 G100 G150 M100 3 G100 G150 LMA LMA LMA CMA CMA CMA LMA*CMA LMA*CMA LMA*CMA 564.211 450.665 178.604 60.677 5.223 21.367 11.116 .479 7.036 0.000 0.000 0.000 0.000 0.041 0.000 0.000 0.631 0.001 Y Y Y Y Y Y Y N Y Source F Value Pr>F Significant

ANOVA analysis of dry tensile strength for Martin Marietta Gravel

Tables 3-13, 3-14, and 3-15 show the results of ANOVA analysis of the dry tensile strength of Marshall 100-mm specimen, ANOVA analysis of the dry tensile strength of Gyratory 100-mm specimen, and ANOVA analysis of the dry tensile strength of Gyratory 150-mm specimen, respectively. Table 3-16 provides the summary and comparison results of ANOVA analysis for different compaction methods for Martin Marietta Gravel. As can be seen from Tables 3-13, 3-14, 3-15 and 3-16, the loose mix aging is the most important factor to influence the dry tensile strength. In addition, compacted mix aging is the second important factor. The above observation is valid for the Marshall 100-mm 46

specimen, Gyratory 100-mm specimen, and Gyratory 150-mm specimen for Martin Marietta Gravel.

Table 3-13 ANOVA Analysis of the Dry Tensile Strength (Marshall 100-mm specimen) Rank 1 2 3 Source LMA LMA * CMA CMA F Value 289.328 15.274 0.204 Pr>F 0 0.001 0.659 Significant Y Y N

Table 3-14 ANOVA Analysis of the Dry Tensile Strength (Gyratory 100-mm specimen) Rank 1 2 3 Source LMA CMA LMA * CMA F Value 193.825 5.223 0 Pr>F 0 0.041 0.998 Significant Y Y N

Table 3-15 ANOVA Analysis of the Dry Tensile Strength (Gyratory 150-mm specimen) Rank 1 2 3 Source LMA CMA LMA * CMA F Value 462.676 21.981 1.952 Pr>F 0 0 0.188 Significant Y Y N

47

Table 3-16 Summarizing comparison ANOVA Analysis for the dry tensile strength Rank Compaction Method M100 1 G100 G150 M100 2 G100 G150 M100 3 G100 G150 Source LMA LMA LMA LMA*CMA CMA CMA CMA LMA * CMA LMA * CMA F Value 289.328 193.825 462.676 15.274 5.223 21.981 0.204 0 1.952 Pr>F 0.000 0.000 0.000 0.001 0.041 0 0.659 0.998 0.188 Significant Y Y Y Y Y Y N N N

Complete factorial ANOVA analysis for dry tensile strength

An analysis of variance (ANOVA) is performed to determine the significance of the various factors and interactions of these factors to the dry tensile strength. Table 3-17 and Table 3-18 show the test results of dry tensile strength at different conditions for limestone and gravel, respectively. The class level information is explained in Table 3-19. Table 3-20 provided a summary of data analysis. Independent variables in the analysis are the source of aggregate (SOURCE), compaction method (COMP), loose mix aging (LMA), compacted mix aging (CMA). The dependent variable is dry tensile strength. The ANOVA analysis indicates that loose mix aging is the most important factor affecting the dry tensile strength. The analysis also shows that both source of aggregate and compaction method have some effects on the dry tensile strength. Figure 3-3 and 3-4 show the dry tensile strength at different conditions for limestone and gravel, respectively. 48

From the Figures 3-3 and 3-4, one can observe that both 4 hours loose mix aging and 16 hours loose mix aging increase the dry tensile strength in all cases. Dry tensile strength with 24 hours compacted mix aging and with 72 hours compacted mix aging are about the same, indicating that compacted mix aging is not an important factor affecting the dry tensile strength. For both aggregate sources, the 150-mm diameter specimens have lower tensile strength than the 100-mm diameter specimens. It should be noted that the loading rate (2-inches per minute) is the same for both sizes of specimens; therefore, the strain rate for 150-mm diameter specimens is 50% lower than that for the 100-mm diameter specimens. A lower loading strain rate usually produces a lower tensile strength. One can also observe that the dry tensile strengths of limestone are larger than the dry tensile strengths of gravel, indicating the influence of the aggregate source.

Table 3-17 Results Summary of Dry Tensile Strength for the Limestone Loose Compacted Mix Mix Aging Aging No Aging 4 Hrs Aging 16 Hrs Aging 24 h 72 h 24 h 72 h 24 h 72 h 1 2 3 4 5 6 Designation No. Condition Marshall 100 90.13 110.25 166.31 193.20 170.96 173.90 Gyratory 100 81.16 83.56 165.67 173.74 160.48 168.27 Gyratory 150 82.94 82.71 127.67 143.53 141.69 171.63 Mean Value for Limestone (psi)

49

Table 3- 18 Results Summary of Dry Tensile Strength for the Gravel Loose Compacted Mix Mix Aging Aging No Aging 4 Hrs Aging 16 Hrs Aging 24 h 72 h 24 h 72 h 24 h 72 h 1 2 3 4 5 6 Designation No. Condition Marshall 100 86.04 91.79 172.10 187.29 133.82 158.98 Gyratory 100 78.96 78.40 151.41 144.61 140.10 147.48 Gyratory 150 58.40 71.57 167.90 146.96 132.01 150.91 Mean Value for Gravel (psi)

Table 3-19 Class level Information for Dry Tensile Strength Parameters Level of Variations 1. No aging 2. 4 hrs@275F 3. 16hrs@140F 1. 24 hrs@ room temp 2. 72 hrs@ room temp 1. Limestone 2. Gravel 1. Marshall Compaction Method (COMP) 2. Gyratory 4 in 3. Gyratory 6 in

Loose Mix Aging

LMA

Compacted Mix Aging (CMA) Source of Aggregate (SOURCE)

50

Table 3-20 ANOVA Analysis of Dry Tensile Strength Dependent Variable: Dry Tensile Strength Mean Factors Square LMA COMP SOURCE CMA SOURCE * LMA SOURCE * CMA COMP * LMA * CMA SOURCE * COMP * LMA COMP * LMA 67919.85 4210.88 2976.16 1213.95 958.95 951.91 724.69 665.98 412.92 1825.43 113.17 79.99 32.63 25.77 25.58 19.48 17.9 11.1 2.05E-62 5.94E-23 2.66E-13 2.34E-07 3.62E-09 3.12E-06 6.74E-11 3.05E-10 4.56E-07 F Sig.

R Squared = .984 (Adjusted R Squared = .976)

51

Dry Tensile Strength

Tensile Strength (psi)
250 200 150 100 50 0 0 1 2 3 4 5 6 7 Marshall 100 Gyratory 100 Gyratory 150

Condition Number

Figure 3-3 Distribution of Dry Tensile Strength for Honey Creek Limestone

Dry Tensile Strength

Tensile Strength (psi)
200 150 Marshall 100 100 50 0 0 1 2 3 4 5 6 7 Gyratory 100 Gyratory 150

Condition Number

Figure 3-4 Distribution of Dry Tensile Strength for Martin Marietta Gravel

52

3.3.2 Freeze-Thaw Tensile Strength:

Freeze-Thaw tensile strength is the indirect tensile strength of conditioned specimen obtained by indirect tensile strength tests. 3.3.2.1 Comparison of three compaction methods The effects of compaction method and sample size on the freeze-thaw tensile strengths are analyzed in this section. Tables 3-21, 3-22, and 3-23 show statistical comparisons of freeze-thaw tensile strength for the Marshall 100-mm HMA specimens, the Gyratory 100-mm HMA specimens, and Gyratory 150-mm HMA specimens, respectively. Based on these tables, some observations are summarized below.
M100 Samples versus G100 Samples

Table 3-21 indicates that the freeze-thaw tensile strength of samples compacted by the M100 compactor is statistically the same as those samples compacted by the G100 compactor for both aggregate sources.
M100 Samples versus G150 Samples

Table 3-22 indicates that the freeze-thaw tensile strength of samples compacted by the M100 compactor is statistically larger than those samples compacted by the G150 compactor for both aggregate sources.
G100 Samples versus G150 Samples

Table 3-23 indicates that the freeze-thaw tensile strength of samples compacted by the G100 compactor is statistically larger than those samples compacted by the G150 compactor for Honey Creek Limestone, but statistically the same for Martin Marietta Gravel.

53

Table 3-21 Statistical Comparison of 100-mm Diameter Marshall Compacted Samples and 100-mm Diameter Superpave Gyratory Compacted Samples Sample conditioning Source Honey Creek Freeze-Thaw Limestone Martin Marietta Gravel Larger* Same*

Smaller*

Table 3-22 Statistical Comparison of 100-mm Diameter Marshall Compacted Samples and 150-mm Diameter Superpave Gyratory Compacted Samples Sample conditioning Source Honey Creek Freeze-Thaw Limestone Martin Marietta Gravel Larger*

Same*

Smaller*

Table 3-23 Statistical Comparison of 100-mm and 150-mm Diameter Superpave Gyratory Compacted Samples Sample conditioning Source Honey Creek Freeze-Thaw Limestone Martin Marietta Gravel 54 Larger*

Same*

Smaller*

3.3.2.2 Freeze-Thaw tensile strength subjected to loose mix aging and compacted mix aging Tables 3-24 and 3-25 show the statistical analysis results for the influence of loose mix aging and compacted mix aging on the freeze-thaw tensile strength for the Honey Creek Limestone. Tables 3-26 and 3-27 show the statistical analysis results for the influences of loose mix aging and compacted mix aging on the freeze-thaw tensile strength for the Martin Marietta Gravel. For Honey Creek Limestone, Table 3-24 indicates that loose mix aging increases the freeze thaw tensile strength in 18 of 18 conditions. Table 3-25 shows that compacted mix aging increases the freeze thaw tensile strength in 10 of 27 possible comparisons. In 17 of 27 possible comparisons, the freeze-thaw tensile strength are same. For Martin Marietta Gravel, Table 3-26 indicates that loose mix aging increases the freeze-thaw tensile strength in 18 of 18 conditions. Table 3-27 shows that compacted mix aging increases the freeze-thaw tensile strength in 11 of 27 possible comparisons. In 16 of 27 comparisons, the freeze-thaw tensile strengths are the same.

55

Table 3-24 Statistical Comparison of Freeze-Thaw Tensile Strength for Mixtures Subjected to Loose Mix Aging (Honey Creek Limestone) Compaction Method Compacted Mix Aging Hrs 24 Marshall 100 72 Saturation Level 50 70 90 50 70 90 50 Gyratory 100 72 24 70 90 50 70 90 50 24 Gyratory 150 72 Marshall Gyratory Total 70 90 50 70 90

Increase*

Same*

Decrease*

6 12 18

56

Table 3-25 Statistical Comparison of Freeze-Thaw Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Honey Creek Limestone) Compaction Method Loose Mix Aging Hrs No aging Saturation Level 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 Increase*

Same*

Decrease*

Marshall 100

4 hrs aging

16 hrs aging

No aging

Gyratory 100

4 hrs aging

16 hrs aging

No aging

Gyratory 150

4 hrs aging

16 hrs aging Marshall Gyratory Total

3 7 10 57

6 11 17

Table 3-26 Statistical Comparison of Freeze-Thaw Tensile Strength for Mixtures Subjected to Loose Mix Aging (Martin Marietta Gravel) Compaction Method Compacted Mix Aging Hrs 24 Marshall 100 72 Saturation Level 50 70 90 50 70 90 50 Gyratory 100 72 24 70 90 50 70 90 50 24 Gyratory 150 72 70 90 50 70 90 Marshall Gyratory Total

Increase*

Same*

Decrease*

6 12 18

58

Table 3-27 Statistical Comparison of Freeze-Thaw Tensile Strength for Mixtures Subjected to Compacted Mix Aging (Martin Marietta Gravel) Compaction Method Loose Mix Aging Hrs No aging Marshall 100 Saturation Level 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 50 70 90 Increase*

Same*

Decrease*

4 hrs aging

16 hrs aging

No aging Gyratory 100

4 hrs aging

16 hrs aging

No aging Gyratory 150

4 hrs aging

16 hrs aging Marshall Gyratory Total

4 7 11

5 11 16

59

3.3.2.3 ANOVA analysis of freeze-thaw conditioned tensile strength

ANOVA analysis of freeze-thaw conditioned tensile strength for Honey Creek

Limestone Tables 3-28, 3-29, and 3-30 show the results of ANOVA analysis of the freeze-thaw tensile strength of Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 150-mm specimen, respectively. Table 3-31 provides the summary and comparison results of ANOVA analysis for the freeze-thaw tensile strength with different compaction methods for Honey Creek Limestone. Tables 3-28, 3-29, 3-30 and 3-31 indicate that the loose mix aging is the most important factor to influence the freeze-thaw conditioned tensile strength and saturation level is the second important factor for all of Marshall 100-mm specimen, Gyratory 100mm specimen, and Gyratory 150-mm specimen, respectively. For Gyratory 100-mm specimen and Gyratory 150-mm specimen, the compacted mix aging is ranked 3, after loose mix aging and saturation level. For Marshall specimen, the compacted mix aging is ranked 5, after loose mix aging, saturation level, and combined effects of loose mix aging and saturation level.

60

Table 3-28 ANOVA Analysis of the Complete Factorial Experiment for the Freeze-Thaw Tensile Strength (Marshall 100 mm specimen) Rank 1 2 3 4 5 6 7 Source LMA SATLEV LMA * SATLEV LMA * CMA CMA LMA * SATLEV * CMA SATLEV * CMA F Value 509.339 94.861 7.393 6.452 .446 .223 .083 Pr>F 0.000 0.000 0.000 .004 .509 .924 .920 Significant Y Y Y Y N N N

Table 3-29 ANOVA Analysis of the Complete Factorial Experiment for the Freeze-Thaw Tensile Strength (Gyratory 100-mm specimen) Rank 1 2 3 4 5 6 7 Source LMA SATLEV CMA LMA * CMA LMA * SATLEV LMA * SATLEV * CMA SATLEV * CMA F Value 1034.555 138.149 16.991 11.560 4.925 .973 .915 Pr>F 0.000 0.000 0.000 .000 .003 .434 .410 Significant Y Y Y Y Y N N

Table 3-30 ANOVA Analysis of the Complete Factorial Experiment for the Freeze-Thaw Tensile Strength (Gyratory 150-mm specimen) Rank 1 2 3 4 5 6 7 Source LMA SATLEV CMA LMA * CMA LMA * SATLEV LMA * SATLEV * CMA SATLEV * CMA F Value 1531.931 213.162 38.926 22.184 5.481 2.002 .936 61 Pr>F 0.000 0.000 0.000 .000 .001 .115 .401 Significant Y Y Y Y Y N N

Table 3-31 Comparison ANOVA Analysis for the F-T Tensile Strength Compaction Method M100 1 G100 G150 M100 2 G100 G150 M100 3 G100 G150 M100 4 G100 G150 M100 5 G100 G150 M100

Rank

Source LMA LMA LMA SATLEV SATLEV SATLEV LMA * SATLEV CMA CMA LMA * CMA LMA * CMA LMA * CMA CMA LMA * SATLEV LMA * SATLEV LMA * SATLEV * CMA LMA * SATLEV * CMA LMA * SATLEV * CMA SATLEV * CMA SATLEV * CMA SATLEV * CMA

F Value 509.339 1034.555 1531.931 94.861 138.149 213.162 7.393 16.991 38.926 6.452 11.560 22.184 .446 4.925 5.481 .223

Pr>F 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 .004 .000 .000 .509 .003 .001 .924

Significant Y Y Y Y Y Y Y Y Y Y Y Y N Y Y N

G100

.973

.434

G150 M100 7 G100 G150

2.002 .083 .915 .936

.115 .920 .410 .401

N N N N

62

 ANOVA analysis of freeze-thaw tensile strength for Martin Marietta Gravel

Tables 3-32, 3-33, and 3-34 show the results of ANOVA analysis of the freeze-thaw conditioned tensile strength of Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 150-mm specimen, respectively. Table 3-35 shows the summary and comparison of ANOVA analysis results for the freeze-thaw tensile strength with different compaction methods for Martin Marietta Gravel. Tables 3-32, 3-33, 3-34 and 3-35 indicate that the loose mix aging is the most important factor to influence the freeze-thaw conditioned tensile strength, while saturation level is the second important factor for all of Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 150mm specimen. For all the Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 150-mm specimen, the compacted mix aging is ranked 3, after the loose mix aging and saturation level.

Table 3-32 ANOVA Analysis of the Freeze-Thaw Conditioned Tensile Strength (Marshall 100-mm specimen) Rank 1 2 3 4 5 6 7 Source LMA SATLEV CMA LMA * SATLEV LMA * CMA SATLEV * CMA LMA * SATLEV * CMA F Value 623.103 14.454 10.729 5.738 1.698 1.124 0.558 Pr>F 0 0.001 0 0.007 0.197 0.336 0.695 Significant Y Y Y Y N N N

63

Table 3-33 ANOVA Analysis of the Freeze-Thaw Conditioned Tensile Strength (Gyratory 100-mm specimen) Rank 1 2 3 4 5 6 7 Source LMA SATLEV CMA LMA * SATLEV LMA * CMA SATLEV * CMA LMA * SATLEV * CMA F Value 385.32 5.08 2.756 2.375 1.816 0.947 0.308 Pr>F 0 0.002 0.077 0.107 0.177 0.337 0.871 Significant Y Y N N N N N

Table 3-34 ANOVA Analysis of the Freeze-Thaw Conditioned Tensile Strength (Gyratory 150-mm specimen) Rank 1 2 3 4 5 6 7 Source LMA SATLEV CMA LMA * SATLEV LMA * CMA SATLEV * CMA LMA * SATLEV * CMA F Value 550.805 39.318 31.363 6.612 6.165 4.207 2.872 Pr>F 0 0 0 0 0.005 0.007 0.07 Significant Y Y Y Y N N N

64

Table 3-35 Comparison ANOVA Analysis for the F-T Tensile Strength Rank Compaction Method M100 1 G100 G150 M100 2 G100 G150 M100 3 G100 G150 M100 4 G100 G150 M100 5 G100 G150 M100 6 G100 G150 M100 Source LMA LMA LMA SATLEV SATLEV SATLEV CMA CMA CMA LMA * SATLEV LMA * SATLEV LMA * SATLEV LMA * CMA LMA * CMA LMA * CMA SATLEV * CMA SATLEV * CMA SATLEV * CMA LMA * SATLEV * CMA LMA * SATLEV * CMA LMA * SATLEV * CMA F Value 623.103 385.32 550.805 14.454 5.08 39.318 10.729 2.756 31.363 5.738 2.375 6.612 1.698 1.816 6.165 1.124 0.947 4.207 0.558 Pr>F 0 0 0 0.001 0.002 0 0 0.077 0 0.007 0.107 0 0.197 0.177 0.005 0.336 0.337 0.007 0.695 Significant Y Y Y Y Y Y Y N Y Y N Y N N N N N N N

G100

0.308

0.871

G150

2.872

0.07

65

 Complete factorial ANOVA analysis for freeze-thaw tensile strength

Table 3-36 and Table 3-37 show the test results of freeze-thaw tensile strength at different conditions for limestone and gravel, respectively. Table 3-38 provides a summary of ANOVA analysis results. Independent variables in the analysis are the source of aggregate (SOURCE), compaction method (COMP), loose mix aging (LMA), compacted mix aging (CMA), and saturation level (SATLEV). The dependent variable is freeze-thaw tensile strength. The ANOVA analysis indicates that loose mix aging is the most important factor affecting the freeze-thaw tensile strength. Source of aggregate, compaction method and saturation level are also important. Compacted mix aging is not an important factor. Figures 3-5 and 3-6 show the freeze-thaw tensile strength at different conditions for limestone and gravel, respectively. The condition numbers are explained in Tables 3-36 and 3-37. From Figures 3-5 and 3-6, one can see that both 4 hours and 16 hours loose mix aging increase the freeze-thaw conditioned tensile strength. Compacted mix aging shows very little influence on the freeze-thaw conditioned tensile strength. For both aggregate sources, the 150mm diameter specimens have lower freeze-thaw conditioned tensile strength than the 100mm diameter specimens. The data also indicates that freeze-thaw tensile strength decreases with the increase of the level of saturation.

66

Table 3-36 Results Summary of Freeze-Thaw Tensile Strength for the Limestone Loose Saturation Compacted Mix Level Aging 50 No Aging 24 h 72 h 70 24 h 72 h 90 24 h 72 h 50 4 hrs Aging 24 h 72 h 70 24 h 72 h 90 24 h 72 h 50 16 hrs Aging 24 h 72 h 70 24 h 72 h 90 24 h 72 h Mix Aging No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Designation Marshall 100 70.51 79.68 60.60 71.60 50.98 65.09 168.94 166.39 141.00 145.02 116.95 116.85 181.08 160.60 133.75 137.65 128.29 97.26 Gyratory 100 72.78 75.04 69.31 62.64 59.02 52.36 148.92 156.18 129.29 139.98 111.26 122.97 134.99 150.46 126.36 136.29 108.41 115.73 Gyratory 150 69.17 69.96 57.39 54.16 51.35 44.30 122.86 131.92 111.80 116.03 94.67 110.22 141.02 153.05 117.65 131.08 102.86 119.56 Condition Mean Value for limestone

67

Table 3-37 Results Summary of Freeze-Thaw Tensile Strength for the Gravel Loose Saturation Compacted Mix Level Aging 50 No Aging 24 h 72 h 70 24 h 72 h 90 24 h 72 h 50 4 hrs Aging 24 h 72 h 70 24 h 72 h 90 24 h 72 h 50 16 hrs Aging 24 h 72 h 70 24 h 72 h 90 24 h 72 h Mix Aging No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Designation Marshall 100 81.86 88.63 70.57 75.25 70.21 66.36 133.80 150.31 142.74 160.76 160.73 159.83 144.18 165.86 157.88 166.13 146.94 146.66 Gyratory 100 71.51 68.12 68.18 64.87 63.88 70.56 138.84 130.41 142.06 140.77 146.00 147.72 154.92 153.73 128.57 138.22 125.99 147.38 Gyratory 150 61.47 65.63 39.78 56.62 45.53 51.12 155.22 121.26 168.19 123.74 162.03 113.01 119.81 144.64 162.10 126.78 137.77 127.12 Condition Mean value for gravel

68

Table 3-38 ANOVA Analysis of Freeze-Thaw Conditioned Tensile Strength Dependent Variable: F-T Conditioned Tensile Strength Source LMA SATLEV SOURCE COMP SOURCE * SATLEV SOURCE * COMP * CMA SOURCE * LMA SOURCE * COMP * LMA * CMA SOURCE * LMA * SATLEV SOURCE * LMA * CMA SOURCE * COMP * LMA SOURCE * CMA COMP * LMA * CMA COMP * CMA LMA * CMA SOURCE * COMP * LMA * SATLEV SOURCE * COMP * SATLEV * CMA CMA LMA * SATLEV Corrected Total R Squared = .975 (Adjusted R Squared = .963) Mean Square 192711.89 7855.23 7756.53 7023.02 5051.03 1244.46 1088.71 1055.82 948.30 780.19 741.01 668.53 599.69 538.84 381.34 341.27 278.30 259.13 253.71 F 3545.35 144.51 142.70 129.20 92.92 22.89 20.03 19.42 17.45 14.35 13.63 12.30 11.03 9.91 7.02 6.28 5.12 4.77 4.67 Sig. 6.9E-166 1.46E-40 1.39E-25 1.25E-37 7.62E-30 9.6E-10 1.05E-08 1.15E-13 2.02E-12 1.4E-06 6.3E-10 0.000551 3.68E-08 7.6E-05 0.001117 2.62E-07 0.000584 0.030083 0.00124

69

F-T Conditioned Tensile Strength for limestone 200

Tensile Strength (psi)

150 100 50 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Condition Designation No. Marshall 100 Gyratory 100 Gyratory 150

Figure 3-5 Distribution of Freeze-Thaw Conditioned Tensile Strength for Limestone

F-T Conditioned Tensile Strength for Gravel 200

Tesnile Strength (psi)

150 100 50 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Condition Designation No. Msrhall 100 Gyratory 100 Gyratory 150

Figure 3-6 Distribution of Freeze-Thaw Conditioned Tensile Strength for Gravel

70

3.3.3 Dry Tensile Strength versus Freeze-Thaw Tensile Strength

The tensile strengths obtained from both conditioned and unconditioned specimens are compared in this section. 3.3.3.1 Comparison of three compaction methods The water conditioning of HMA samples by vacuum saturation, soaking, and a freezethaw cycle normally decreases the tensile strength. Statistical comparisons have been made among groups of sample dry (without conditioning) and after water conditioning with vacuum saturation and a freeze-thaw cycle. This section provides information about the effect of compaction method and sample size on the comparison results between dry tensile strength and freeze-thaw tensile strength. The variable included in this analysis is mixture source (Honey Creek Limestone Vs. Martin Marietta Gravel). Table 3-39 shows statistical comparisons of samples for Honey Creek Limestone, while Table 3-40 shows statistical comparisons of samples for Martin Marietta Gravel. Based on analysis presented in these two tables, the following observations can be made.

For Honey Creek Limestone

1. M100 Samples Table 3-39 indicates that the freeze-thaw tensile strengths are statistically the same as the dry tensile strengths for 2 of 18 conditions. In 16 of 18 comparisons, the freeze thaw tensile strengths are statistically lower than the dry tensile strengths.

71

2. G100 and G150 Samples Table 3-39 indicates that the freeze-thaw tensile strengths are statistically the same as the dry tensile strengths for 2 of 36 conditions. In 34 of 36 comparisons, the freeze-thaw tensile strengths are statistically lower than the dry tensile strengths.

For Martin Marietta Gravel

1. M100 Samples Table 3-40 indicates that the freeze-thaw tensile strengths are statistically the same as the dry tensile strengths for 3 of 9 conditions. In 6 of 9 comparisons, the freeze-thaw tensile strengths are statistically lower than the dry tensile strengths.

2. G100 and G150 Samples Table 3-40 indicates that the freeze-thaw tensile strengths were statistically the same as the dry tensile strengths for 17 of 36 conditions. In 17 of 36 comparisons, the freeze-thaw tensile strengths are statistically lower than the dry tensile strengths. In 2 of 36 comparisons, the freeze-thaw tensile strengths are statistically larger than the dry tensile strengths.

72

Table 3-39 Statistical Comparison of Dry Tensile Strength and Freeze-Thaw Tensile Strength (Honey Creek Limestone) Compaction Method Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 Marshall 100 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 24 72 2 2 4 1 1 1 1 3 3 2 3 2 3 3 3 3 3 3 3 3 3 2 3 2 3 16 34 50 Increase* Same* Decrease*

No aging

73

Table 3-40 Statistical Comparison of Dry Tensile Strength and Freeze-Thaw Tensile Strength (Martin Marietta Gravel) Compacted Compaction Method Loose Mix Mix Aging Aging Hrs Hrs No aging Marshall 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 24 72 3 2 5 1 2 1 6 17 23 2 9 17 26 1 3 2 1 3 1 1 3 3 2 2 1 1 3 3 3 3 1 1 1 2 2 2 3 Increase* Same* Decrease*

74

3.3.3.2 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength In theory the conditioned tensile strength is expected to be lower than the dry tensile strength, because conditioning in water according to the AASHTO T 283 can lead to moisture damage. Figures 3-7, 3-8, and 3-9 show the comparisons of the dry tensile strength versus freeze-thaw tensile strength at different conditions for Honey Creek Limestone. Figures 3-10, 3-11, and 3-12 show the comparisons of the dry tensile strength versus freeze-thaw tensile strength at different conditions for Martin Marietta Gravel. The condition numbers indicated in these figures are defined in Tables 3-36 and 3-37 for Honey Creek Limestone and Martin Marietta Gravel, respectively. The results show that freeze-thaw tensile strengths are lower than the dry tensile strengths for both Honey Creek Limestone and Martin Marietta Gravel. For all the HMA specimens, including Marshall 100-mm specimen, Gyratory 100-mm specimen and Gyratory 150-mm specimen, the same trend is observed in these figures. It is clear that freeze-thaw conditioning is crucial in the conditioning process. One freeze-thaw cycle should be included in the proposed water stripping test procedure.

75

Marshall 100 for Limestone 250

Tensile Strength (psi)

200 150 100 50 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Condition Designation No. Dry F-T Conditioned

Figure 3-7 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Honey Creek Limestone (Marshall 100-mm Diameter Specimen)

Gyratory 100 for Limestone 200 150 100 50 0 1 2 3 4 5 6 7 8 9 101112131415161718 Condition Designation No. Dry F-T Conditioned

Strength for Honey Creek Limestone (Gyratory100-mm Diameter Specimen) 76

Tesnile Strength (psi)

Figure 3-8 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile

Gyratory 150 for Limestone 200 150 100 50 0 1 2 3 4 5 6 7 8 9 101112131415161718 Condition Designation No.
Figure 3-9 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Honey Creek Limestone (Gyratory 150-mm Diameter Specimen)

Tensile Strength (psi)

Dry F-T Conditioned

Marshall 100 for gravel 200

Tensile Strength (psi)

150 100 50 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Dry F-T Conditioned

Condition Designation No.

Figure 3-10 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Martin Marietta Gravel (Marshall 100-mm Diameter Specimen) 77

Gyratory 100 for Gravel 200

Tensile Strength (psi)

150 100 50 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Condition Designation No.

Dry F-T conditioned

Figure 3-11 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Martin Marietta Gravel (Gyratory100-mm Diameter Specimen)

Gyratory 150 for Gravel 200

Tensile Strength (psi)

150 100 50 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Condition Designation No.

Dry F-T Conditioned

Figure 3-12 Comparisons of Dry Tensile Strength and Freeze-Thaw Tensile Strength for Martin Marietta Gravel (Gyratory 150-mm Diameter Specimen)

78

3.3.4 Level of Saturation

The experimental program includes saturation level at 50, 70, and 90 percent for both Honey Creek Limestone and Martin Marietta Gravel. Tables 3-41, 3-42, and 3-43 show the statistical comparison results of the influence of level of saturation on freeze-thaw tensile strength for the Honey Creek Limestone. Tables 3-44, 3-45, and 3-46 are results for Martin Marietta Gravel. For both Honey Creek Limestone and Martin Marietta Gravel, the freeze-thaw tensile strength decreases with the increase of saturation level in most cases. From the ANOVA analysis (Table 3-38) for the freeze-thaw tensile strength, the saturation level is an important factor. Figure 3-13 shows the relationship of freeze-thaw conditioned tensile strength with saturation level for limestone at the 24 hrs compacted mix aging. The results show that freeze-thaw tensile strength decreases with the increase of saturation level in all cases. A good correlation exists between freeze-thaw conditioned tensile strength and saturation level. For the 72 hrs compacted mix aging, the similar trend of relationship is obtained. The test results also indicate that freeze-thaw tensile strength of HMA specimens saturated to the lower end of range may be significantly different than those of the same HMA specimens saturated to the upper end for a similar air void content. As a result, the TSR values could also be significantly different since the dry tensile strength is the same.

79

Table 3-41 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 70% vs. 50% (Honey Creek Limestone) Compaction Method Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 Marshall 100 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 24 72 0 2 2

Increase*

Same*

Decrease*

No aging

6 10 16

80

Table 3-42 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 50% (Honey Creek Limestone) Compaction Method Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 Marshall 100 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 24 72 0 0 0

Increase*

Same*

Decrease*

No aging

6 12 18

81

Table 3-43 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 70% (Honey Creek Limestone) Compaction Method Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 Marshall 100 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 24 72

Increase*

Same*

Decrease*

No aging

0 2 2

6 10 16

82

Table 3-44 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 70% vs. 50% (Martin Marietta Gravel) Compaction Method Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 Marshall 100 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 1 1 24 72

Increase*

Same*

Decrease*

No aging

4 8 12

2 3 5

83

Table 3-45 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 50% (Martin Marietta Gravel) Compaction Method Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 Marshall 100 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 1 24 72 1

Increase*

Same*

Decrease*

No aging

2 9 11

3 3 6

84

Table 3-46 Statistical Comparison of Freeze-Thaw Tensile Strength subjected to different Levels of Saturation: 90% vs. 70% (Martin Marietta Gravel) Compaction Method Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 Marshall 100 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 100 24 72 4 hrs aging 24 72 16 hrs aging 24 72 No aging Gyratory 150 24 72 4 hrs aging 24 72 16 hrs aging Marshall Gyratory Total 1 24 72 1

Increase*

Same*

Decrease*

No aging

3 11 14

2 1 3

85

F-T Conditioned Tensile Strength Vs. Saturation level

200.00 180.00 160.00 M100, No LMA, 24 hrs CMA G100, No LMA, 24 hrs CMA G150, No LMA, 24 hrs CMA M100, 4hrs LMA, 24 hrs CMA G100, 4hrs LMA, 24 hrs CMA G150, 4hrs LMA, 24 hrs CMA M100, 16hrs LMA, 24 hrs CMA G100, 16hrs LMA, 24 hrs CMA G150, 16hrs LMA, 24 hrs CMA

Tensile Strength (psi)

140.00 120.00 100.00 80.00 60.00 40.00 20.00 0.00

40%

50%

60%

70%

80%

90%

100%

Saturation Level

Figure 3-13 Relationship of the F-T Tensile Strength Versus. Saturation Level

3.3.5 Water Sensitivity

Tensile strength ratios of 70 to 80 percent are typically used as acceptance levels for the AASHTO T 283 method of test. Tensile strength ratios from portions of this study are shown in Tables 3-47, 3-48, and 3-49. Table 3-47 shows the sources of the materials, type of compaction, and conditioning associated with 70 and 80 percent minimum tensile strength ratios for this portion of the study. The mixtures prepared with the Honey Greek Limestone and Martin Marietta Gravel would pass the 70 and 80 criteria for the most conditions. For Honey Greek Limestone, almost all the tensile strength ratio passed the 80% at the saturation of 50%. At the saturation of 70%, 6 of 14 conditions passed the TSR of 70%, while 8 of 14 passed 80%. At the saturation of 90%, just half of these conditions can pass the TSR of 70%, the other tensile strength ratios are below the TSR of 70%. 86

For Martin Marietta Gravel, at the saturation of 50%, almost all the tensile strength ratio passed the TSR of 80%. The same observation can be made at the saturation of 70%. At the saturation of 90%, 4 of 18 conditions passed the TSR of 70%, 14 of 18 passed TSR of 80%. Tables 3-48 and 3-49 also show that Marshall 100-mm specimen fails the TSR of 80% at high saturation level while Gyratory 150-mm specimen pass TSR of 80% at the same saturation level for both sources of aggregates. This indicates that Marshall 100-mm specimen has a lower tensile strength ratio than the Gyratory 150-mm specimen at the same saturation level.

Table 3-47 Acceptable Mixtures Mixtures Honey Greek F-T Limestone Martin Marietta F-T Gravel 70*, 80 70*, 80 70*, 80 M100 G100 G150 Conditioning Method of Compaction

87

Table 3-48 Tensile Strength Ratio for Freeze-Thaw Conditioning (Honey Greek Limestone) Loose Compaction Method Mix Aging Hrs No aging Marshall 100 4 hrs aging 16 hrs aging No aging Gyratory 100 4 hrs aging 16 hrs aging No aging Gyratory 150 4 hrs aging 16 hrs aging Marshall Gyratory Total 2 Compacted Mix Aging Hrs 24 72 24 72 24 72 24 72 24 72 24 72 24 72 24 72 24 72 2 70%

Saturation Level 50 80% 70% 70 80% 70% 90 80%

4 12 16

2 4 6

2 6 8

2 5 7

0 0 0

*Note: * Meet 70% or 80% retained tensile strength ratio.

88

Table 3-49 Tensile Strength Ratio for Freeze-Thaw Conditioning (Martin Marietta Gravel) Loose Compaction Method Mix Aging Hrs No aging Marshall 100 4 hrs aging 16 hrs aging No aging Gyratory 100 4 hrs aging 16 hrs aging No aging Gyratory 150 4 hrs aging 16 hrs aging Marshall Gyratory Total 1 Compacted Mix Aging Hrs 24 72 24 72 24 72 24 72 24 72 24 72 24 72 24 72 24 72 1

Saturation Level 50 70% 80%

70 70% 80%

90 70% 80%

5 12 17

0 1 1

6 10 16

1 3 4

5 9 14

*Note: * Meet 70% or 80% retained tensile strength ratio.

89

3.3.6 Tensile Strength Ratio

Table 3-50 and Table 3-51 show the results of TSR at different conditions for Honey Creek Limestone and Martin Marietta Gravel, respectively. Table 3-52 shows the ANOVA analysis of the tensile strength ratio from the complete factorial experiment for Honey Creek Limestone,. Table 3-53 shows the ANOVA analysis of the tensile strength ratio from complete factorial experiment for Martin Marietta Gravel. The results show that the loose mix aging, saturation level and compaction method are important factors influencing the tensile strength ratio. Compacted mix aging ranks behind these three dominant factors. Figure 3-14 shows Tensile Strength Ratio of Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 100-mm specimen at different conditions for Honey Creek Limestone. The figure shows that Tensile Strength Ratio decreases with the saturation level for all the Marshall 100-mm, Gyratory 100-mm, and gyratory 150-mm specimens. The TSR value is higher for the specimens with 4 hours loose mix aging and 16 hours loose mix aging at the condition when other testing parameters are the same. Figure 3-15 shows Tensile Strength Ratio of Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 100-mm specimen at different conditions for Martin Marietta Gravel. The figure shows that Tensile Strength Ratio changes a little with the different compacted mix aging. The TSR value is higher for the specimens with 4 hours loose mix aging than the specimen with 16 hours loose mix aging when other test parameters are the same.

90

Table 3-50 Comparison of TSR of M100, G100 and G150 at Different Conditions (Honey Greek Limestone) Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 No aging 24 72 24 72 24 72 4 hrs aging 24 72 24 72 24 72 16 hrs aging 24 72 24 72 Saturation Level 50% 50% 70% 70% 90% 90% 50% 50% 70% 70% 90% 90% 50% 50% 70% 70% 90% 90% Condition Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 M100 0.782 0.723 0.672 0.649 0.566 0.590 1.016 0.861 0.848 0.751 0.703 0.605 0.945 0.896 0.813 0.741 0.704 0.653 TSR G100 0.897 0.898 0.854 0.750 0.727 0.627 0.899 0.899 0.780 0.806 0.672 0.708 0.841 0.894 0.787 0.810 0.676 0.688 G150 0.834 0.846 0.692 0.655 0.619 0.536 0.962 0.919 0.876 0.808 0.742 0.768 0.995 0.892 0.830 0.764 0.726 0.697

91

Table 3-51 Comparison of TSR of M100, G100 and G150 at Different Conditions (Martin Marietta Gravel) Loose Mix Aging Hrs Compacted Mix Aging Hrs 24 72 No aging 24 72 24 72 24 72 4 hrs aging 24 72 24 72 24 72 16 hrs aging 24 72 24 72 Saturation Level 50% 50% 70% 70% 90% 90% 50% 50% 70% 70% 90% 90% 50% 50% 70% 70% 90% 90% Condition Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 M100 0.950 0.970 0.820 0.860 0.820 0.760 0.777 0.803 0.829 0.858 0.934 0.853 1.239 0.907 1.241 0.993 1.096 0.924 TSR G100 0.906 0.869 0.863 0.827 0.809 0.900 0.917 0.902 0.938 0.973 0.964 1.021 1.106 1.042 0.918 0.937 0.899 0.999 G150 1.053 0.917 0.681 0.791 0.780 0.714 0.924 0.825 1.002 0.842 0.965 0.769 0.908 0.958 1.228 0.840 1.044 0.842

92

Table 3-52 ANOVA Analysis of the Tensile Strength Ratio for Limestone Dependent Variable: TSR Source SATLEV LMA CMA COMP * LMA COMP COMP * LMA * CMA COMP * CMA COMP * LMA * SATLEV COMP * SATLEV Mean Square 0.67 0.14 0.06 0.04 0.02 0.01 0.01 0.00 0.00 F 415.97 87.56 36.57 23.74 15.52 7.89 7.39 2.59 1.85 Sig. 1.81E-51 2.51E-23 2.15E-08 4.21E-14 1.19E-06 1.28E-05 0.000982 0.01252 0.123704

R Squared = .922 (Adjusted R Squared = .883)

Table 3-53 ANOVA Analysis of the Tensile Strength Ratio for Gravel Dependent Variable: TSR Source LMA COMP * CMA LMA * SATLEV COMP * LMA * CMA COMP * LMA SATLEV COMP * LMA * SATLEV COMP CMA Error Mean Square 0.36 0.14 0.06 0.05 0.05 0.03 0.02 0.02 0.00 0.00 F 91.43 36.17 16.39 13.88 11.63 8.42 6.23 5.85 0.66 Sig. 5.84E-24 9.48E-13 1.6E-10 3.61E-09 6.83E-08 0.0004 1.31E-06 0.003888 0.419461

R Squared = .845 (Adjusted R Squared = .770)

93

1.1 1 0.9 TSR 0.8 0.7 0.6 0.5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Condition No.

M100 G100 G150

Figure 3-14 TSR distributions of M100, G100 and G150 at different conditions (Honey Creek Limestone)

1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

TSR

M100 G100 G150

Condition No.

Figure 3-15 TSR distributions of M100, G100 and G150 at different conditions (Martin Marietta Gravel)

94

3.4 Summary and Conclusions Based on the test performed and data analysis presented in this chapter, the following observations can be made. 1. Loose mix aging is the most important factor affecting the dry tensile strength. Both sources of aggregate and compaction method have an effect on the dry tensile strength. Compacted mix aging is not as an important factor as the above mentioned factors. 2. Loose mix aging is the most important factor affecting the freeze-thaw tensile strength. Source of aggregate, compaction method, and saturation level are also important to freeze-thaw tensile strength. Compacted mix aging is not as an important factor as the above mentioned factors. 3. Freeze-thaw tensile strengths are lower than the dry tensile strengths for both limestone and gravel, indicating that freeze-thaw cycle is crucial in the conditioning process. One freeze-thaw cycle should be included in the proposed water stripping test procedure. 4. Freeze-thaw tensile strength decreases with the increase of saturation level. Based on Tables 3-50 and 3-51, Marshall 100-mm specimen has a lower tensile strength ratio than the Gyratory 150-mm specimen at the same saturation level. The saturation level could be increased to 80%-90% for the Gyratory 150-mm, based on the analysis of test results presented in this chapter, in order for the TSR values to be even closer between Marshall 100-mm specimens and Gyratory 150-mm specimens.

95

5. Loose mix aging, saturation level, and compaction method are important factors influencing the tensile strength ratio. Compacted mix aging ranks behind these three dominant factors in affecting the tensile strength ratio.

96

CHAPTER IV DEVELOPMENT OF STRIPPING TEST PROCEDURURE FOR SUPERPAVE SPECIMENS

4.1 Introduction Based on the analysis of various factors affecting the stripping test results from the complete factorial experimental test program presented in Chapter III, regression analysis technique is used to evaluate the relationship of the tensile strength ratio among Marshall 100-mm specimen, Gyratory 100-mm specimen, and Gyratory 150-mm specimen. Based on regression analysis results, the proposed sample conditioning procedure is developed. 4.2 Analysis of Tensile Strength Ratio Figures 4-1 and 4-2 show the TSR regression simulation between Marshall 100-mm specimen and Gyratory 100-mm specimen for Honey Creek Limestone and Martin Marietta Gravel, respectively. Figures 4-3 and 4-4 show the TSR regression simulation between Marshall 100-mm specimen and Gyratory 150-mm specimen for Honey Creek Limestone and Martin Marietta Gravel, respectively. The tensile strength ratios of Gyratory 150-mm compacted samples are similar to the tensile strength ratios of Marshall 100-mm compacted samples. The R-square is 0.83 for

97

Honey Creek Limestone and 0.63 for Martin Marietta Gravel, respectively. On the contrary, no relationship can be found between Gyratory 100-mm specimen and Marshall 100-mm specimen, since the R-square is 0.478 for Honey Creek Limestone and 0.011 for Martin Marietta Gravel, respectively. It is recommended that Gyratory 150-mm compacted samples should be used in the proposed procedure. There is no relationship between Gyratory 100-mm specimen and Marshall 100-mm specimen because of different compaction methods. Due to the difference in size, the compaction method has a greater influence on the Gyratory 100-mm specimen than it has on the Gyratory 150-mm specimen. This could be the reason why the Gyratory 150-mm specimen has a closer relationship with the Marshall 100-mm specimen than the Gyratory 100-mm specimen. In previous analysis presented in Chapter III, (i.e., Table 3-52 and Table 3-53), it was found that the loose mix aging, saturation level, and compaction method are three important factors affecting the tensile strength ratio. The values of TSR are different for specimens conditioned with different loose mix aging procedure. Therefore, loose mix aging is required in the proposed procedure. The TSR values of Marshall 100-mm specimen at standard loose mix aging (16hrs) condition are compared with the TSR values of Gyratory 150-mm specimen at different loose mix aging conditions for Honey Creek Limestone and Martin Marietta Gravel in Figures 4-5 and 4-6, respectively. It is seen that the TSR values of Gyratory 150mm specimen at four hours loose mix aging are about the same as the TSR values of Marshall 100mm specimen at standard loose mix aging condition for both Honey Creek Limestone and Martin Marietta Gravel. The four hours loose mix aging for Gyratory 150-mm specimens is similar to sixteen hours loose mix aging on Marshall 100-mm specimen. It is recommended that the four hours loose 98

mix aging for Gyratory 150-mm compacted samples should be used in the proposed procedure. Based on the analysis of dry tensile strength, freeze-thaw conditioned tensile strength and tensile strength ratio for both Honey Creek Limestone and Martin Marietta Gravel, compacted loose mix aging is not an important factor for the stripping test. Dry tensile strength, freeze-thaw conditioned tensile strength and tensile strength ratio are not significantly changed by different compacted mix aging procedure, (i.e., Table 3-20, Table 3-38, Table 3-52, and Table 3-53) between 24 hours compacted mix aging and 72 hours compacted mix aging. It is recommended that 24 hours compacted mix aging to be used in the proposed test procedures. Linear regression analysis is performed to investigate the relationships between the TSR values and the degree of saturation. The results are summarized in Table 4-1 and plotted in Figure 4-7. The slopes and intercepts for the regression lines are presented in the table along with the R-square values. The R-square values ranged from 0.85 to 0.99, suggesting a strong correlation between the TSR values and the level of saturation. All the regression lines have relatively low but negative slopes, indicating a reduction in TSR values with an increase of degree of saturation. TSR values of specimens saturated to the lower end of range are significantly higher than the same specimens saturated to the upper end of range. Figure 4-8 shows a comparison of TSR values between Marshall 100-mm specimens with 16 hours aging and Gyratory 150-mm specimens with 4 hours aging at different saturation level. The TSR values of G150 are a little higher than the M100 at the same saturation level. It is recommended that the saturation level be increased to a range of 80%-90% in the proposed test procedure. 99

Based on the ANOVA analysis for both Honey Creek Limestone and Martin Marietta Gravel presented in Chapter III, one cycle freeze-thaw condition is an important factor that needs to be included in the proposed test procedures. The proposed test procedure is shown in Table 4-2, in which the suggested conditioning procedure is marked.
1.0

.9

.8

.7

G100

.6 .5 .6 .7 .8 .9 1.0 1.1

Rsq = 0.4785

M100

Figure 4-1 TSR Regression Simulations for Honey Creek Limestone (M100 Vs. G100)
1.1

1.0

.9

G100

.8 .7 .8 .9 1.0

Rsq = 0.0111

M100

Figure 4-2 TSR Regression Simulations for Martin Marietta Gravel (M100 Vs. G100) 100

1.1

1.0

.9

.8

.7

.6

G150

.5 .5 .6 .7 .8 .9 1.0 1.1

Rsq = 0.8339

M100

Figure 4-3 TSR Regression Simulations for Honey Creek Limestone (M100 Vs. G150)

1.0

.9

.8

.7

G150

.6 .6 .7 .8 .9 1.0

Rsq = 0.6308

M100

Figure 4-4 TSR Regression Simulations for Martin Marietta Gravel (M100 Vs. G150)

101

1.2 1 0.8 TSR 0.6 0.4 0.2 0

(24,50%) (24,70%) (24,90%)

M100 with Standard LMA G150 with No LMA G150 with 4 hrs LMA G150 with 16 hrs LMA

Figure 4-5 Comparisons of TSR Values of M100 at Standard Loose Mix Aging Condition with G150 at different Loose Mix Aging Condition (Honey Creek Limestone)

1.4 1.2 1 TSR 0.8 0.6 0.4 0.2 0 (24,50%) (24,70%) (24,90%) M100 with Standard LMA G150 with No LMA G150 with 4 hrs LMA G150 with 16 hrs LMA

Figure 4-6 Comparisons of TSR Values of M100 at Standard Loose Mix Aging Condition with G150 at different Loose Mix Aging Condition (Martin Marietta Gravel) 102

1.1 1 0.9

0.8 0.7 0.6 0.5 0.4 30% 40% 50% 60% 70% 80% 90% 100%

Saturation Level

M100, No LMA, 24 hrs CMA M100, 4hrs LMA, 24 hrs CMA M100, 16 hrs LMA, 24 hrs CMA G150, No LMA, 24 hrs CMA G150, 4 hrs LMA, 24 hrs CMA G150, 16 hrs LMA, 24 hrs CMA G100, No LMA, 24 hrs CMA G100, 4 hrs LMA, 24 hrs CMA G100, 16 hrs LMA, 24 hrs CMA

TSR %

Figure 4-7 Relationship of TSR and Saturation Level for Honey Creek Limestone

1.1 1 0.9

TSR

0.8 0.7 0.6 0.5 30%

Marshall 100 Gyratory 150

50%

70%

90%

110%

Saturation Level

Figure 4-8 TSR Values of M100 and G150 with Different Saturation Level

103

Table 4-1 Summary of Regression between TSR values and the Degree of Saturation Sample Type Loose Mix Aging No M100 4 16 No G100 4 16 No G150 4 16 Slope -0.005 -0.008 -0.006 -0.005 -0.006 -0.007 -0.004 -0.006 -0.004 Intercept 1.053 1.403 1.242 1.091 1.246 1.322 1.123 1.181 1.508 R-square 0.99 0.996 0.993 0.933 0.97 0.96 0.85 0.99 0.921

4.3 Partial Factorial Test Program The partial factorial test program for the Traprock and Stocker Sand is depicted in Figure 4-9. The purpose of the partial factorial program is two fold: (a) to validate the conclusions and proposed test procedures using the two new aggregate sources, and (b) to validate the applicability of the test procedure for polymer modified binders.

104

Table 4-2 Summary of the Suggested Test Procedure for Gyratory 150-mm (G150) Specimen

Factors Investigated

Level of Variations 1. No aging

Suggested

Loose Mix Aging

2. 4 hrs@275F 3. 16hrs@140F 1. 24 hrs@ room temp

Compacted Mix Aging 2. 72 hrs@ room temp 1. None Freeze-Thaw Cycle 2. One 1. Marshall Compaction Method 2. Gyratory 4 in 3. Gyratory 6 in 1. 50% Saturation Level 2. 70% 3. 90%

105

Figure 4-9 Partial Factorial Experimental Test Program

106

4.4 The Partial Factorial Test Results and Validation of Proposed Test Procedure The complete mix design details are presented in Appendix for two aggregates sources using polymer modified binder PG 70-22. Table 4-3 shows comparison of the values of Tensile Strength Ratio between Marshall 100-mm specimen using standard AASHTO T283 procedure and Gyratory 150-mm specimen using the proposed procedure for both Traprock Source and Stocker Sand Gravel. The mean and standard deviation of tensile strength ratio values for the Marshall 100mm specimen using standard procedure and for Gyratory 150-mm specimen using the proposed procedure are shown in Table 4-3. The mean value of Gyratory 150-mm specimen using the proposed procedure is almost the same as the Marshall 100-mm specimen using standard procedure for both sources of aggregate. Therefore, the test results of the partial factorial test program validate the proposed test procedure. The TSR values of Traprock source and Stocker Sand and gravel are higher than that of Honey Creek Limestone and Martin Marietta Gravel. This is because the modified asphalt binder is used in the partial factorial experimental test, which improved the moisture damage resistance for the specimens.

107

Table 4-3 Comparison of TSR of M100 at standard procedure and G150 at proposed procedure Loose Mix Aging 16 hrs 24 h Aging Traprock 4 hrs 24 h Aging 90% 70% TSR Value Compacted Mix Aging Saturation Level Sample Number 1 2 3 1 2 3 16 hrs 24 h Stocker Sand Gravel 4 hrs 24 h Aging 90% Aging 70% 1 2 3 1 2 3 1.023 0.883 0.859 0.945 0.893 0.995 0.944 0.051 0.922 0.089 Marshall Gyratory 100 0.829 0.879 1.080 0.909 1.052 0.870 0.944 0.096 0.929 0.133 150 Mean Standard deviation Statistics

Aggregate Source

108

4.5 Conclusions and Recommendations Based on the analysis of test data conducted on the HMA specimens prepared using four different aggregate sources and two different asphalt binders, it can be concluded that aggregate source, method of compaction, specimen size, loose mix aging, compacted HMA aging, freeze-thaw conditioning, and saturation level can exert influences on the outcome of test results in terms of dry tensile strength, freeze-thaw conditioned tensile strength, and tensile strength ratio (TSR). Based on the premise that TSR for the Superpave HMA specimens be compatible with TSR for the Marshall HMA specimens under current AASHTO T 283, the key steps in the proposed test procedure for moisture induced damage for Superpave HMA specimens include the following.
Use of Gyratory 150 mm specimen is recommended as the specimen size for

water stripping test.

After mixing the aggregate and asphalt binder, age the loose HMA for 4 hours at

275 F.
Heat the loose mixtures to the required compaction temperature before

compacting.
After compacting the loose mixture to the air-void content of 7 percent 1

percent, the HMA samples are extruded from the compaction mold and allowed to aging for 24 hours at room temperature.
Place the HMA samples into water and saturate the samples to a saturation level

between 80 and 90 percent.

109

 Condition the sample for a freeze cycle (16 hr at 0 F) and a thaw-soak cycle (24

hr at 140 F).
Put the sample into the water bath at room temperature for 2 hrs prior to

commencing the indirect tension test.

110

CHAPTER V HMA RUT RESISTANCE CHARACTERIZED BY SIMPLE MATERIAL PROPERTIES

5.1 Introduction and Objectives of the Chapter Rutting is a primary failure mode of hot mix asphalt pavements. Since the mid-1980s, modified asphalt concrete mixture has been widely used to minimize rutting failures. Since a lot of research was based on the unmodified asphalt, more research about the modified asphalt was needed. Rutting predictions by physical properties of HMA and mechanics properties of HMA are desirable. They can be used to screen and differentiate rutting properties. The objective of this research is to investigate the relationship between the rutting performance of the modified HMA and their physical or strength properties. Research Plan and Experimental Data A structured laboratory test is used to investigate the rutting properties of hot mixture asphalt.
5.2.1 Research Approach

The research approach is depicted in Figure 5-1, where four binders are used to develop HMA specimens. The rut depth is measured from the APA tests. The measured

111

mechanical properties of HMA include indirect tensile strength and unconfined compressive strength. Other physical properties of the mixture are measured from the optimum asphalt mix design. The experimental data is from previous research by Jadoun, F. M. (MSc. Thesis, 2001).

SBR B Unmodified Asphalt PG 58-28 Modified Asphalt SBS A SBS K

Binder Content Stability Flow

Indirect Tensile Strength Strength Properties

Mixture Characteristics

VMA VFA Unit Weight

Unconfined Compressive Strength

APA Test Correlate with

Rutting Depth

Figure 5-1 Framework of the Research Approach

112

5.2.2 Materials

The source of aggregates came from local aggregate products. It is a blending of two sources. The coarse aggregates (retaining on sieve #4 and above) were limestone brought from Akron Crushed Limestone Company in Akron, Ohio. The fine aggregates (passing #4) were sand gravel brought from Allied Corporation Inc. in Massillon, Ohio. Aggregate gradation used was according to ODOT requirements for heavy traffic (Type-1H). Figure 5-2 shows the gradation used in JFM. In this study, two types of polymer modifiers were used. The SBS (Styrene Butadiene Styrene) provided by Ashland and Koch and the SBR (Styrene Butadiene Rubber) provided by BASF.
Aggregate Gradation Curves of the Upper and Lower Limits of the ODOT Type-1H and the Job Mix Formula Used in this Research
120

100

80

Percen Passing

Lower Limit

60

Upper Limit JMF

40

20

0 No. 200 No. 100 No. 50 No. 30 No. 16 No. 8 No. 4 3/8" 1/2" 3/4"

Sieve Size

Figure 5-2 ODOT Requirements for Heavy Traffic (Type-1H)

113

5.2.3 Mixture Designs and Physical Properties

All of the mixtures were designed using the Marshall Mixture design procedure. Table 5-1 provided a summary of the mixture properties of the HMA samples. The numerical values were the average results of three samples.

Table 5-1 Optimum Properties of Marshall Design for All Mixture

Binder Mixture Type Stability Content (%) Unmodified PG58-28 2% SBR Butanol + PG58-28 3% SBR Butanol + PG58-28 2% SBS Ashland + PG58-28 3% SBS Ashland + PG58-28 4% SBS Ashland + PG58-28 5% SBS Ashland + PG58-28 2% SBS Koch + PG58-28 3% SBS Koch + PG58-28 4% SBS Koch + PG58-28 5% SBS Koch + PG58-28 5.92 6.20 6.30 5.80 5.82 5.75 5.92 5.82 5.92 5.93 6.23 2265 2100 2200 2622 2620 2405 2600 2548 2550 2890 2700 (lb)

Flow (0.01 in) 7.60 9.10 11.35 8.55 9.50 9.80 9.27 8.73 9.37 10.10 10.90

VMA (%) 13.40 13.97 13.75 13.30 13.52 13.56 13.65 13.57 13.56 13.52 14.24

Unit VFA (%) Weight (pcf) 72.5 74.5 75.5 73.0 73.0 74.5 73.0 74.5 73.0 73.0 74.5 145.4 144.8 145.0 145.4 145.1 145.0 145.0 145.0 145.2 145.2 144.5

114

5.2.4 Asphalt Mixture Test and Mechanical Strength Properties

The parameters including indirect tensile strength and unconfined compression strength were used to correlate to the rutting properties. 5.2.4.1 Introduction of asphalt mixture test
Indirect Tensile Strength Test

The indirect tensile strength test was conducted according to the ASTM D 4123 standards. The indirect tensile strength was calculated by the following equation:

T
Where: T S =tensile strength,

= (2000 P) / td

(Eq. 5.1)

P =maximum load carried by the sample, t =specimen thickness,

d =specimen diameter

Unconfined Compression Test

The unconfined compression test was conducted to obtain the stress-strain data of HMA specimens. The specimen dimensions were 4 inch diameter and 8 inch height specimens. The reason for using this height is to minimize the frictional end effect in order to get more accurate results for the measured vertical strains. The test was conducted at 104F temperature. 5.2.4.2 Mechanical strength properties test results Indirect tensile test and unconfined compression test were investigated in since they were simple to be conducted and usually regarded as good predictors by some researchers. The detailed test results are given in the Table 5-2.

115

Table 5-2 Indirect Tensile Strength Test Results and Unconfined Compression Test Results for All Mixture

Unconfined Indirect Tensile Mixture Type Strength (psi) Strength (psi) Unmodified PG58-28 2% SBR Butanol + PG58-28 3% SBR Butanol + PG58-28 2% SBS Ashland + PG58-28 3% SBS Ashland + PG58-28 4% SBS Ashland + PG58-28 5% SBS Ashland + PG58-28 2% SBS Koch + PG58-28 3% SBS Koch + PG58-28 4% SBS Koch + PG58-28 5% SBS Koch + PG58-28 105 118.7 103.7 114.6 126.4 113.8 116.6 107.6 134.2 136.4 128.4 34 37.5 45.5 56.5 64 70 69 67.5 81.5 78 65.5 Compressive

116

5.2.5 Rutting Test Data from APA Test

The Asphalt Pavement Analyzer developed by the Pavement Technology Inc. was used to obtain the rutting data of HMA specimens under the simulated wheel loads. 5.2.5.1 Introduction of asphalt pavement analyzer test (APA Test) The APA applied a 100 lb aluminum wheel onto a pneumatic linear host pressurized to 100 psi, which rests on the test specimen. The rut depth is measured at 8000 cycles of wheel load passes. 5.2.5.2 APA test data results Table 5.3 Summarizes the APA rut depth values for all the test specimens at different modified polymer content. Five rut depth reading were taken, then testing started and reading were taken after 500, 1000, 2000, 4000, and 8000 cycles of wheel passes. Table 5-3 Asphalt Pavement Analyzer Test Results for All Mixture Rut Depth (mm) Mixture Type Unmodified PG58-28 2% SBR Butanol + PG58-28 3% SBR Butanol + PG58-28 2% SBS Ashland + PG58-28 3% SBS Ashland + PG58-28 4% SBS Ashland + PG58-28 500 cycles 4.44 0.96 1.47 1.36 1.90 1.03 1000 cycles 5.31 1.40 1.87 2.06 2.47 1.19 2000 cycles 6.13 2.02 2.52 3.26 3.20 1.58 4000 cycles 7.19 2.50 3.80 3.77 3.89 2.37 8000 cycles 8.60 3.97 5.17 4.94 4.95 3.37

117

5.3 Analysis of Test Results In general, the higher the binder grade, the stiffer the binder, and the greater the rutting resistance. The low rutting resistance was connected with high asphalt content. The volumetric properties of the asphalt mixture were believed to affect rutting, such as void in the mineral aggregate (VMA) and void filled with asphalt (VFA). By providing appropriate VMA, the rutting could be minimized, and durability would be enhanced. The use of polymer modified asphalt was believed to be another factor to affect the rutting resistance. The strength-based laboratory test methods such as indirect tensile test, creep test, unconfined compression test, Hamburg wheel test, asphalt pavement analyzer have been used to evaluate rutting of asphalt concrete mixes.

5.3.1 Single Parameter Linear Regression Analysis

Use the Physical Mixture Properties to Correlate to Rut Depth

A linear regression analysis was conducted to investigate the relationship between asphalt concrete mixture properties (Including asphalt content, air void, VMA, VFA, stability flow, unit weight and polymer modified asphalt) and rutting performance. Table 5-1 provided a summary of the mixture properties of the HMA samples. The test results of rut depth measured by Asphalt Pavement Analyzer Test were shown in the Table 5-2. Separate regression analysis was conducted for the properties obtained from laboratory specimens. A summary of the analysis results was given in Table 5-4. The highest correlation to the rutting depth was the Marshall stability with R-square value of 0.396. It showed a poor correlation. The study of rutting by Huber and Heiman also found that 118

Marshall stability was a poor indicator of rutting performance. From the summary of the results, no strong correlation was found between the physical mixture properties and the rutting performance. Because the binder contributes approximately 20 percent to rut resistance, while remaining 80 percent is dominated by the aggregate structure. One parameter is not enough to describe the rutting behavior, since it is a complicated phenomenon. Asphalt concrete mixtures contained modifiers significantly improve the rutting resistance compared to the unmodified asphalt concrete mixtures. Figure 5-3, 5-4 showed that both 2%, 3% and 4%, 5% modifier contents improved the rutting resistance over the unmodified asphalt concrete mixtures for all the conditions. At 2 and 3 percent modifier content, SBR Butanol provided better resistance to rutting than the SBS Koch before 2000 cycles. After which, SBS Koch performed better. Furthermore, the slope of the SBS Koch curve was flatter than the other modifies curves, meaning that SBS Koch modified mixture did not accumulate rut depths with time. The same thing was found at 4 and 5 percent modifier content. Figure 5-5 was plotted to show the rutting depth versus modifier content at the 8000 number of wheel load cycles. Based on the literature review, many researchers analyzed the APA rut depth results by incorporating the rut depth value accumulated after 8000 load cycles in their analysis procedures. Therefore, the APA rut depth value developed after 8000 wheel load passes was used in this research. Figure 5-3 showed us that the rut depth decreased steeply with the increase of the polymer content before 4 percent, after which, the rut depth changed a little. In general, it can be seen that the best rutting resistance can be achieved at a polymer content of 3 to 4 percent. The figure also showed that SBS Koch modified mixtures were experiencing less rutting depths than all other modified mixtures. SBA 119

Ashland and SBR Butanol modified mixture showed rut depth reduction of 55 percent and 45 percent respectively. Figures 5-5 showed the rut depth versus modifier content at the number of 8000 wheel load cycles.

Table 5-4 Summary of Regression Analysis Result between Rut Depth and Mixture Properties Mix parameter Binder Content Stability
Flow VMA VFA Unit Weight

Slope -0.993 -0.006 -1.102 -3.803 -0.476 4.799

Intercept 9.619 18.537 14.144 55.564 38.77 -689.571

R-square 0.007 0.396 0.277 0.207 0.045 0.311

10
Rut Depth (mm)

8 6 4 2 0 0 2000 4000 6000 8000 10000 Number of Load Cycles

Unmodified 2% SBR B 2% SBS A 2% SBS K 3% SBR B 3% SBS A 3% SBS K

Figure 5-3 Rutting Depth Vs. Number of Cycles for 2% and 3% Modifiers 120

10
Rut Depth (mm)

8 6 4 2 0 0 2000 4000 6000 8000 10000 Number of Load Cycles

Unmodified 4% SBS A 4% SBS K 5% SBS A 5% SBS K

Figure 5-4 Rutting Depth Vs. Number of Cycles for 4% and 5% Modifiers

Rutting Depth (mm)

10 9 8 7 6 5 4 3 2 1 0
0% 2%

8000 Cycles

SBR B SBR A SBS K

3% Modifier Content

4%

5%

Figure 5-5 Rutting Depth Vs. Modifiers Content After 8000 cycles
Use the Strength Properties to Correlate to Rut Depth

The analysis above showed the importance of developing a simple laboratory test that can relate to the rutting performance. Physical mixture properties contribute to the rutting 121

resistance; however, they cannot be used to predict the rut depth. The fundamental strength and deformation properties of the asphalt concrete mixtures are believed to provide the useful information to relate to the rutting performance. Strength tests, besides being simpler and easier to carry out compared to stiffness tests, fully mobilize the resistance of both the asphalt binder and the aggregate structure comprising asphalt concrete, and are thus a better method for evaluating resistance to permanent deformation. Indirect tensile test and unconfined compression test were investigated in this research since they were simple to be conducted and usually regarded as good predictors by some researchers. The detailed test results are given in the Table 5-2. Poor correlations were found for both the vertical displacement at maximum load and the indirect tensile strength (R2-value=0.392). The correlation figure between rut depth and indirect tensile strength was shown in the Figure 5-6. Although, the indirect tensile strength was used as an indicator to predict rutting by some researchers, it was not surprising. The similar correlation was found by Brown and Cross. The correlation for the indirect tensile strength was found almost the same as the Marshall stability (R2-value=0.396). The Marshall stability didnt account for plastic flow of the materials under slow or prolonged loading, though it had been used to predict rutting in the past. A better correlation was found between the unconfined compressive strength and rut depth (R2-value=0.563). A plot of the regression was shown in the Figure 5-6 and Figure 5-7. Figure 5-8 and Figure 5-9 showed the influence of polymer adding to the asphalt to the indirect tensile strength and unconfined compressive strength respectively. It was clear that polymer increased both indirect tensile strength and unconfined compressive strength. It worked the best at a modifier content range of 3 to 4 percent. 122

10 9 8 7

Rut Depth (mm)

6 5 4 3 2 1 0 100 110 120 130 140 Rsq = 0.3920

Indirect Tensile strength (psi)

Figure 5-6 Regression Simulations for Indirect Tensile Strength

10

Rut Depth (mm)

0 30 40 50 60 70 80 90

Rsq = 0.5631

Compression Strength (psi)

Figure 5-7 Regression Simulations for Unconfined Compressive Strength 123

160
Indirect Tensile Strength (psi)

150 140 130 120 110 100 90

0% 2% 3% 4%

SBR B SBR A SBS K

5%

Modifier Content

Figure 5-8 Indirect Tensile Strength Vs. Modified Content for all Mixture

110
Compressive Strength (psi)

100 90 80 70 60 50 40 30 20
0% 2% 3% 4%

SBR B SBR A SBS K

5%

Modifier Content

Figure 5-9 Compressive Strength Vs. Modified Content for all Mixture

124

5.3.2 Multiple Parameter Linear Regression Analysis

Since mixture properties and the strength parameters were found influencing rut depth, all the variables including air void, stability, unit weight, flow, void in mineral aggregate, and void filled with asphalt, indirect tensile strength and unconfined compressive strength were used to relate to rutting performance. Stepwise regression method was used to determine which variables have the greatest significance on laboratory rutting performance. From the correlation test results, unconfined compressive strength was the most significant predictor variable. In fact, when unconfined compressive strength alone was used to predict rut depth at 8000 cycles in the APA, the R2-value is 0.563. The VMA was an important variable too, which also was found by Stacy in his research. Form the analysis of variables correlation matrix, we found that the strongest correlation with rut depth was unconfined compressive strength with a correlation in excess of 0.75. Correlation was negative, meaning that as the value of one variable went down, the value of the other variable tended to go up. The variables stability and indirect tensile strength were also correlated with rut depth, while the strong correlations were also found between unconfined compressive strength and indirect tensile strength, stability which meant multicollinearity existing. Multicollinearity problems can occur when one or more variables are linear related to the other variables in the model. Since many of the predictor variables were statistically related to each other in some way. SPSS formed the model without multicollinearity problem, which showed in Table 5-5. The first one used unconfined compressive strength variable alone in the model with the R-square value 0.563. When VMA was added to the model, the R-square value increased to 0.842. While adding the other variables, the R-square value would not increase anymore, which 125

indicated these two variables have provided the optimum R-square value. The significance values from the Table 5-6 for the model 2 was 0.001 (<0.05), which indicated the model was significant. As a result, the VMA from the mixture properties and the unconfined compressive strength from the strength properties were included in the model.

Rutting Depth = 70.96 - 0.113 unconfined compressive strength - 4.429 VMA (Equ. 5.2)

Figure 5-10 is a plot of the measured using APA test and predicted values of rut depth used model, the line of equality demonstrated the equation does a reasonable job of prediction. Table 5-5 Linear Regression Analysis Results for Model Summary Std. Error of the Estimate R Square Change 1 2 .750(a) .918(b) .563 .842 .515 .802 1.541573 .983527 .563 .279 F Change 11.599 14.110 Sig. F Change .008 .006 Change Statistics

Model

R Square

Adjusted R Square

df1

df2

1 1

9 8

126

Table 5-6 Significance of the Model

Model Regression 1 Residual Total Regression 2 Residual Total

Sum of Squares 27.563 21.388 48.951 41.213 7.739 48.951

df 1 9 10 2 8 10

Mean Square 27.563 2.376

F 11.599

Sig. .008(a)

20.606 .967

21.302

.001(b)

12
Predicted Rut Depth (mm)

10 8 6 4 2 0 0 2 4 6 8 10 12 Measured Rut Depth (mm)

Figure 5-10 Predicted versus Measured Rutting Potential

127

5.4 Conclusions An experimental test results conducted by Jadoun, F. M. provided pertinent data for investigating the influence of polymer modifiers on the properties of HMA specimens, particularly rutting resistance. Since many mixture properties and strength characteristics have been pointed out in the literature that would relate to the rut performance, regression analysis was utilized to quantify these relationships for the modified asphalt concrete mixture. Based on the analysis results, the VMA and unconfined compressive strength are used in the prediction model. Because rutting is a complicated phenomenon, some researchers believe that the binder contributes approximately 20 percent to rut resistance, while remaining 80 percent is dominated by the HMAs structure. In this prediction model, VMA is related to the properties of asphalt binder, while, unconfined compressive strength represents characteristics of HMAs structure. This is why the model can predict the rutting behavior.

Adding polymers to virgin binder enhanced the mechanical properties of the HMA including both indirect tensile strength and unconfined compressive strength. Modifier content, by weight of 3 to 4 percent appears to provide best results.

According to the APA test results, asphalt concrete mixtures containing modifiers exhibit significantly improved rutting resistance, compared to the unmodified asphalt concrete mixtures.

Rut depth decreased steeply with the increase of the polymer content up to 4 percent. Modifier content of greater than 4% does not to seem to offer any 128

additional benefits in enhancing rutting resistance. An optimum polymer content of 3 to 4 percent is observed.

From the summary of the results, no strong correlation was found between the physical mixture properties and the rutting performance. Physical mixture properties cannot explain the rutting performance observed from APA tests.

Poor correlation (R2-value=0.392) was obtained between the indirect tensile strength and the rut depth. A better correlation (R2-value=0.563) was found between the unconfined compressive strength and rut depth.

The study using the stepwise regression method showed that the rut depth can be statistically related to unconfined compressive strength and VMA with R-square of 0.842.

129

CHAPTER VI RELATIOSHIP INVESTIGATION BETWEEN LOW TEMPERATURE CRACKING AND SIMPLE LABORATORY TEST

6.1 Introduction Thermal cracking of asphalt concrete pavements is a serious problem in many regions of the United States. Thermal cracking is attributed to tensile stresses induced in asphalt concrete pavement by temperature drops to extremely low levels. If the pavement is cooled to a low temperature, tensile stresses develop as the pavement contracts. The friction between the pavement and the base layer resist the contraction. When the tensile stress induced in the pavement equals the strength of asphalt concrete mixture at that temperature, a microcrack develops at the edge and surface of the pavement. At colder temperature or repeated temperature cycles, the crack penetrates the full depth and width of the asphalt concrete layer. Polymer-modified asphalt concrete mixture has been widely used to decrease low temperature cracking failures of flexible pavement, however, the research is limited. There is a need to investigate the HMAs fundamental mechanical property that can be linked back to HMAs low temperature cracking behavior.

130

There are two objectives in this research. The first one is to explore the influence of polymer modifiers adding to the asphalt binders and mixtures to modify their properties, particularly low temperature cracking resistance. The second objective is to investigate the modified asphalt concrete mixtures fundamental mechanical property that can be linked back to HMAs low temperature cracking behavior. In this research, three polymer modifiers are used in the laboratory test. The influences of the modifiers to thermal cracking resistance are investigated. Three parameters obtained from the indirect tensile test, resilient modulus test, and dynamic modulus test are correlated to low temperature cracking properties obtained from the TSRST test. The experimental data is from previous research by Jadoun, F. M. (MSc. Thesis, 2001). 6.2 Experimental Tests
Thermal Stress Restrained Specimen Test

TSRST system developed is an automatic, closed-loop system specifically designed to measure the tensile stress in an asphalt concrete specimen that is cooled at a constant rate while restrained form contraction. The test system consists of a load system, data acquisition system, temperature control system, and specimen alignment stands. The load, data acquisition, and temperature control system are controlled with a personal computer. The load system includes a load frame, a step motor and two swivel jigs. The load frame consists of two aluminum base plates that are supported by four connecting circular rods. A step motor is mounted on the top base plate and a load cell is attached to the bottom base plate. The height of the load frame can be adjusted to accommodate various specimen lengths. Figure 6-1 shows the load system in details. 131

The alignment of the specimen is critical to obtaining meaningful test results. Poor alignment may cause bending stresses in the specimen. The stand provides concentric and perpendicular alignment between the platens and the specimen using an epoxy compound and secure the specimen and platens while the epoxy sets. The stand is mounted vertically. The Figure 6-2 shows the schematic of specimen alignment stand. The Thermal Stress Restrained Specimen Test is conducted according to AASHTO Designation TP10. This standard is based on the Strategic Highway Research program (SHRP) product 1021. As the air temperature slowly decreases and approaches freezing temperature, the asphalt binder layer in the pavement structure begins to contract; however, continuous length of asphalt concrete layer and the friction between asphalt concrete and base layer resist the contraction. As the temperature drops, the stress continues to increase until it is equal to the tensile strength of asphalt concrete. As the point, the asphalt concrete should fracture or crack. The validity of this theory has been studied extensively in the field, mostly in Canada and northern regions of America. The TSRST attempts to simulate these field conditions. As asphalt concrete specimen is cooled at a specific rate (10 degrees Celsius per hour) starting at or slightly above the freezing temperature, as the specimen begins to contract, instrumentation senses the contraction and loads are applied in the opposite direction to stretch the specimen back to its original length. So effectively, the specimen is restrained from contracting, the temperature continues to decrease and just like in the field, tensile stress increases until the specimen fractures.

132

Figure 6-1 Schematic of TSRST apparatus

Figure 6-2 Schematic of specimen alignment stand 133

 Indirect Tensile Strength Test

The indirect tensile strength test was conducted according to the ASTM D 4123 standards. A four-inch diameter Marshall specimen was loaded in compression along the vertical diameter plane at a fixed deformation rate of 2 inch/min. until failure occurred. This loading configuration develops a relatively uniform tensile stress perpendicular to the direction of the applied load and along the vertical diameter plane, which ultimately caused the specimen to fail by splitting along the vertical diameter. A wide curved strip of 0.5 inch width was used to provide a uniform loading width which produces a uniform stress distribution.

Resilient Modulus Test

The resilient modulus test is the most common test for measuring the stiffness modulus of HMA. The test setup is similar to that of indirect tensile strength test. For the resilient modulus test, the recommended load range according to the ASTM D 4123 standards is that to induce 10 to 50 percent of the tensile strength. Load is typically applied for 0.1 second, along with a rest period of 0.9 seconds.

Dynamic Modulus Test

The dynamic modulus can be defined as the absolute value of the complex modulus that defines the elastic properties of a linear viscoelastic material subjected to sinusoidal loading. The dynamic modulus test is used to measure the stress-strain relationship in a pavement structure under a repeated load. The dynamic modulus test was conducted according to the ASTM D 3497 standards. 134

6.3 Analysis of Test Results Table 6-1 summarized the TSRST results for all modified mixture, each number shown is an average of three specimens. Figure 6-3 shows the temperatures at fracture for the different polymer modified mixture mixtures and the unmodified as well. It can be clearly seen that all the modified mixtures have improved the fracture temperature by decreasing it. SBS Ashland was the best in lowering the fracture temperature by 17 percent compared the unmodified. SBS Koch modified mixture show 11 percent improvement in fracture temperature. On the other hand, SBR modified mixtures have also shown improvement in fracture temperature by 6 percent for both SBR Goodyear and SBR Butanol modified mixture. No clear trend can be observed for the optimum modified content. Figure 6-4 illustrates the tensile strength for the different polymer modified mixture mixtures and the unmodified as well. It shows clearly that SBS Koch modified mixtures have increased the tensile strength by 73 percent. SBS Ashland increased the tensile strength by 30 percent. SBR Goodyear has improved about 25 percent, while SBR Butanol has increased the tensile strength of the unmodified by 21 percent. It can be concluded that SBS modifiers perform better in improving the tensile strength and the fracture temperature than the SBR modifiers. SBS Koch modifier was the best in improving the tensile strength by an average of 73 percent; whereas, SBS Ashland modifier showed the best in improving the fracture temperature with and average of 17 percent. In general, for maximum improvement on fracture temperature and tensile strength, 4 percent modifier seems to be the optimum.

135

Statistical analysis using regression techniques was used to determine the level of correlation between the measured laboratory mixture responses to low temperature cracking. Three simple laboratory tests including indirect tensile test, resilient modulus test, and dynamic modulus were used in the data analysis. Table 6-2 summarized indirect tensile strength, resilient modulus, and dynamic modulus test results for all modified mixture, each number shown is an average of three specimens. Figures 6-5, 6-6, and 6-7 showed the relationship between TSRST fracture strength and indirect tensile test, resilient modulus test, and dynamic modulus respectively. According to the R-square values, it can be seen that only indirect tensile strength has a fair correlation (R2value=0.58) to the measured TSRST fracture strength, the resilient modulus (R2value=0.21) and dynamic modulus (R2-value=0.20) did not provide a good definitive relationship with TSRST fracture strength. Figures 6-8, 6-9, and 6-10 showed the relationship between TSRST fracture temperature and indirect tensile test, resilient modulus test, and dynamic modulus respectively. The parameters obtained from the indirect tensile test (R2-value=0.17), resilient modulus test (R2-value=0.05), and dynamic modulus test (R2-value=0.27) have no relationships with TSRST fracture temperature. The parameters obtained from the HMAs simple laboratory test are not enough to evaluate the HMAs low temperature cracking properties. The thermal cracking properties of HMA are more related to the asphalt binder properties and polymer modifiers. More asphalt binder tests should be conducted. The properties of the asphalt binder should be considered in evaluating the characteristics of HMAs low temperature cracking.

136

Table 6-1 TSRST Test results for All Mixture Mixture Type At Fracture Temperature (C) Unmodified PG58-28 2% SBR Goodyear+ PG58-28 3% SBR Goodyear+ PG58-28 4% SBR Goodyear+ PG58-28 5% SBR Goodyear+ PG58-28 2% SBR Butanol + PG58-28 3% SBR Butanol + PG58-28 4% SBR Butanol + PG58-28 2% SBS Ashland + PG58-28 3% SBS Ashland + PG58-28 4% SBS Ashland + PG58-28 5% SBS Ashland + PG58-28 2% SBS Koch + PG58-28 3% SBS Koch + PG58-28 4% SBS Koch + PG58-28 5% SBS Koch + PG58-28 -24.8 -25.7 -28.2 -25.6 -26.1 -26.9 -25.0 -27.3 -29.4 -27.0 -31.8 -28.3 -26.5 -29.1 -28.1 -26.6 Tensile Strength (psi) 285.2 335.7 392.1 322.8 379.9 374.3 300.7 363.7 358.7 369.8 361.7 387.1 467.4 499.3 558.7 445.8

137

Table 6-2 Indirect Tensile Strength, Resilient Modulus, and Dynamic Modulus Test Results for All Mixture Indirect Tensile Mixture Type Strength (psi) Unmodified PG58-28 2% SBR Goodyear+ PG58-28 3% SBR Goodyear+ PG58-28 4% SBR Goodyear+ PG58-28 5% SBR Goodyear+ PG58-28 2% SBR Butanol + PG58-28 3% SBR Butanol + PG58-28 2% SBS Ashland + PG58-28 3% SBS Ashland + PG58-28 4% SBS Ashland + PG58-28 5% SBS Ashland + PG58-28 2% SBS Koch + PG58-28 3% SBS Koch + PG58-28 4% SBS Koch + PG58-28 5% SBS Koch + PG58-28 105 103.4 108.7 102.2 105.3 118.7 103.7 114.6 126.4 113.8 116.6 107.6 134.2 136.4 128.4 Modulus (psi) 554349 628089 563584 567097 511445 735138 621271 638129 585152 483574 583001 754924 648145 654621 750771 Modulus (psi) 234000 321000 254000 261000 266000 315000 334000 322000 312000 359000 368000 335000 348000 346000 320000 Resilient Dynamic

138

45 40 Fracture Temperature (-C) 35 30 25 20 15 10 5 0

SBR Goodyear SBR Butanol SBS Ashland SBS Koch

0%

2%

3% Modifier Content

4%

5%

Figure 6-3 Fracture Temperature Vs. Modified Content in TSRST

600 Fracture Tensile Strength (psi)

500

SBR Goodyear SBR Butanol SBS Ashland SBS Koch

400

300

200

100 0% 2% 3% Modifier Content 4% 5%

Figure 6-4 Fracture Tensile Strength Vs. Modified Content in TSRST

139

600

TARST Fracture Strength (psi)

500

400

300

200 100 110 120 130 140

Rsq = 0.5859

Indirect Tensile Strength (psi)

Figure 6-5 Regression between TSRST Fracture Strength and Indirect Tensile Strength

600

TARST Fracture Strength (psi)

500

400

300

200 400000 500000 600000 700000 800000

Rsq = 0.2143

Resilient Modulus (psi)

Figure 6-6 Regression between TSRST Fracture Strength and Resilient Modulus 140

600

TARST Fracture Strength (psi)

500

400

300

200 220000 260000 300000 340000 380000

Rsq = 0.2097

Dynamic Modulus (psi)

Figure 6-7 Regression between TSRST Fracture Strength and Dynamic Modulus
-24 -25

STRST Fracture Temperature (C)

-26 -27 -28 -29 -30 -31 -32 100 110 120 130 140

Rsq = 0.1796

Indirect Tensile Strength (psi)

Figure 6-8 Regression between TSRST Fracture Temperature and Indirect Tensile Strength 141

-24 -25

STRST Fracture Temperature (C)

-26 -27 -28 -29 -30 -31 -32 400000 500000 600000 700000 800000

Rsq = 0.0533

Resilient Modulus (psi)

Figure 6-9 Regression between TSRST Fracture Temperature and Resilient Modulus
-24 -25 -26

STRST Fracture Temperature (C)

-27 -28 -29 -30 -31 -32 220000 260000 300000 340000 380000

Rsq = 0.2778

Dynamic Modulus (psi)

Figure 6-10 Regression between TSRST Fracture Temperature and Dynamic Modulus 142

6.4 Conclusions and Recommendations: 1. SBS modifiers perform better than the SBR modifiers in improving the tensile strength in the Thermal Stress Restrained Specimen Test. SBS Koch modified

mixtures have shown the best in improving the fracture temperature. Maximum improvement in fracture temperature and the tensile strength improvement is achieved at 4 percent modifier content. 2. From statistical analysis using regression techniques, only indirect tensile strength has a fair correlation to the measured TSRST fracture strength, the resilient modulus and dynamic modulus did not provide a good definitive relationship with TSRST fracture strength. The parameters obtained from the indirect tensile test, resilient modulus test, and dynamic modulus test have no relationships with TSRST fracture temperature. 3. The parameters obtained from the HMAs simple laboratory test is not enough to evaluate the HMAs low temperature cracking properties, the properties of the asphalt should be considered.

143

CHAPTER VII CONCLUSIONS AND RECOMMENDATIONS

7.1 Summary of Moisture Damage Conclusions and Recommendations A structured laboratory test program was conducted and the proposed procedure was given in this study.

7.1.1 Summary of the Moisture Damage Work Completed

A structured laboratory test program was conducted to investigate the effects of the various factors on the HMA specimens resistance to moisture damage. The experimental test program includes a complete factorial and a partial factorial experimental program. The factors investigated include aggregate sources (Honey Creek Limestone and Martin Marietta Gravel), compaction methods (Marshall and Gyratory), specimen size (4 inch and 6 inch), loose mix aging methods (none versus 16 hrs @ 140 F, as well as 4 hrs @ 275 F), saturation level (55, 75, and 90%), and freeze-thaw cycle (none versus one F/T cycle). Honey Creek Limestone and Martin Marietta Gravel are mixed with virgin asphalt binder PG 64-22 to prepare HMA specimens. The partial factorial experimental program is reduced based on the conclusions from the analysis of the complete factorial experimental program. The factors investigated in the partial factorial

144

experiment tests include aggregate sources (Ontario Traprock Source and Stocker Sand and Gravel), compaction methods (Marshall and Gyratory), specimen size (4 inch and 6 inch), loose mix aging methods (AASHTO T283, 16 hrs @ 140 F versus Superpave specifications, 4 hrs @ 275 F), saturation level (90%), and freeze-thaw cycle (one freeze/thaw cycle @ 16 hrs @ 0 F). Ontario Traprock Source and Stocker Sand and Gravel are aggregate sources which are mixed with the modified asphalt binder PG 70-22 to prepare HMA specimens for the partial factorial experimental program. The APPENDIX B at the end of the thesis provides the test results of the Marshall and Gyratory mix design for four sources of aggregate. APPENDIX A provides the test results of the complete factorial experimental program including dry tensile strength, freeze-thaw conditioned tensile strength, and tensile strength ratio. The partial factorial experimental test results are presented in Chapter IV. In the data analysis, the results of the complete factorial experimental program are useful in identifying and evaluating the contributions of the individual factors, along with all the possible interactions among the various factors. The results of the partial factorial experimental program, on the other hand, are useful for validating findings from the complete factorial experimental results. In this research, the effects of different factors on dry tensile strength, freeze-thaw conditioned tensile strength, and tensile strength ratio (TSR) are investigated. The indicator of moisture damage susceptibility is the TSR. Thus, the variability of test results is statistically examined via. mean, standard deviation, and coefficient of variation (COV). Analysis based on Pair-wise comparisons is used to compare the impact of two individual factors while maintaining all other factors at a constant level. The ANOVA 145

(Analysis of Variance) technique is used to perform F-tests (Type III S: partial sum sequences) to identify the contribution of the main test variables and their interactions. Regression analysis technique is used to (1) evaluate the tensile strength ratio relationship between different compaction methods, and (2) evaluate the relationship between tensile strength ratio and saturation level. The effects of all the factors investigated in this research are summarized in this research. The recommendations and proposed test procedure are given based on the test results and the accompanied analysis.

7.1.2 Conclusions and Recommendations

Based on the analysis of test data conducted on the HMA specimens prepared using four different aggregate sources and two different asphalt binders, it can be concluded that aggregate source, method of compaction, specimen size, loose mix aging, compacted HMA aging, freeze-thaw conditioning, and saturation level can exert influences on the outcome of test results in terms of dry tensile strength, freeze-thaw conditioned tensile strength, and tensile strength ratio (TSR). Based on the premise that TSR for the Superpave HMA specimens be compatible with TSR for the Marshall HMA specimens under current AASHTO T 283, the key steps in the proposed test procedure for moisture induced damage for Superpave HMA specimens include the following. 1. Use of Gyratory 150 mm specimen is recommended as the specimen size for water stripping test. 2. After mixing the aggregate and asphalt binder, age the loose HMA for 4 hours at 275 F. 146

3. Heat the loose mixtures to the required compaction temperature before compacting. 4. After compacting the loose mixture to the air-void content of 7 percent 1 percent, the HMA samples are extruded from the compaction mold and allowed to aging for 24 hours at room temperature. 5. Place the HMA samples into water and saturate the samples to a saturation level between 80 and 90 percent. 6. Condition the sample for a freeze cycle (16 hr at 0 F) and a thaw-soak cycle (24 hr at 140 F). 7. Put the sample into the water bath at room temperature for 2 hrs prior to commencing the indirect tension test.

7.1.3 Implementation Recommendations

The recommended test procedure for determining the moisture damage potential of Superpave HMA is established based on the premise that TSR produced by the new procedure with the 6 inch Superpave specimens would be compatible with TSR for the 4 inch Marshall specimens. Thus, it is important to keep in mind that current study is assuming the applicability of the AASHTO T283 for the 4 inch Marshall specimens. Implementation of the proposed test procedure for Superpave HMA specimens requires that ODOT engineers continue to monitor field performance of the same HMA specimens tested in the lab. In addition, establishing minimum tensile strength and tensile strength ratio values based on traffic level, allowable rutting depth, and expected fatigue life is a key issue that needs to be further addressed for full implementation of the

147

moisture induced damage test as an integral part of ODOT Superpave HMA design protocol. The implementation of the research findings in the form of modification of the pertinent section of ODOT Construction and Materials Specifications may result in benefits, including more confidence in the mix design for resisting moisture induced damage. In addition, the proposed test procedure shortens the test duration due to a reduction of aging time to four (4) hours for the loose mix.

7.1.4 Secondary Observations of Experimental Results

This section provides a brief summary of secondary observations, as they are not the main objective of the present study. It should also be cautioned that these observations should not be generalized or interpreted outside the context that these are secondary observations. 7.1.4.1 Aggregate source The aggregate source has an effect on the dry tensile strength, freeze-thaw tensile strength, and tensile strength ratio based on the ANOVA analysis presented in chapter IV. The limestone has higher dry tensile strength than the gravel. For the F-T conditioned tensile strength, the limestone is similar to the gravel. However, when the modified asphalt binder is used in compacting the HMA specimens, the influence of aggregate source on the tensile strength ratio decreases. This observation is in contradiction with ODOT past experiences. More aggregates sources need to be tested before this observation can be further substantiated. 148

7.1.4.2 Method of compaction The compaction method has an effect on the dry tensile strength, freeze-thaw tensile strength, and tensile strength ratio based on the ANOVA analysis presented in chapter IV. No tensile strength ratio relationship can be found between the Marshall 100-mm compacted specimen and Gyratory 100-mm specimen. This indicates that in order to test Gyratory compacted HMA for moisture damage resistance, different Gyratory specimen size (150-mm) may be needed to produce similar results as the Marshall 100-mm specimens. 7.1.4.3 Specimen size Gyratory 150-mm specimen is recommended to be used in the proposed test procedure. The tensile strength ratio of Gyratory 150-mm compacted specimen exhibits a strong correlation with the tensile strength ratio of Marshall 100-mm compacted specimen. 7.1.4.4 Loose mix aging method The loose mix aging is the most important factor for the dry tensile strength, freezethaw conditioned tensile strength and tensile strength ratio. TSR values of Gyratory 150mm specimen at 4 hours loose mix aging are about the same as the TSR values of Marshall 100-mm specimen at standard loose mix aging condition for both limestone and gravel. Therefore, 4 hours loose mix aging is recommended in the proposed test procedure based on the comparisons with the 16 hours aging.

149

7.1.4.5 Compacted mix aging method The compacted mix aging is not an important factor for the dry tensile strength, freezethaw conditioned tensile strength, and tensile strength ratio. Therefore, 24 hours of compacted mix aging could be used in the proposed test procedure, since dry tensile strength, freeze-thaw conditioned tensile strength and tensile strength ratio with 24 hours compacted mix aging are statistically the same as those with 72 hours compacted mix aging. 7.1.4.6 Conditioning procedure The freeze-thaw conditioning is an important factor for the freeze-thaw tensile strength, and tensile strength ratio. Freeze-thaw conditioned tensile strengths are lower than the dry tensile strengths for both limestone and gravel, indicating that freeze-thaw cycle is crucial in the conditioning process. One freeze-thaw cycle should be included in the proposed test procedure. 7.1.4.7 Saturation level The saturation level is an important factor for the freeze-thaw tensile strength and tensile strength ratio. Freeze-thaw conditioned tensile strength decreases with the increases of saturation level. From Figures 3-14 and 3-15, it can be seen that some specimens at 50% saturation level pass TSR requirement, but fail at a 70% or 90% level of saturation, even though 50% saturation is within the AASHTO T 283(99) specifications. Figure 4-8 provides a validation that the saturation level for the Gyratory compacted specimens should be increased. It is recommended that the saturation level be increased to 80%-90% in the proposed test procedure. 150

7.1.5 Recommendations for Future Study

1. Effects of polymer modified asphalt binder and anti-stripping additives on the moisture resistance of the various HMA are encouraged to be completely investigated, since they are an economical and effective way to improve the HMAs resistance to the moisture damage.

2. It is important to incorporate indirect tensile strength as a design and evaluation tool for Superpave mixture and establish minimum tensile strength and tensile strength ratio based on traffic level, allowable rutting depth, and expected fatigue life. The minimum tensile strength criteria could be used along with the tensile strength ratio values as a part of the Superpave mix design criteria. There is a need to set up a simple, reasonable and dependable method for mix design and performance evaluation system for Superpave mixtures.

7.2 Summary of the Rutting and Thermal Cracking Conclusions and Recommendations An experimental test results conducted by Jadoun, F. M. provided pertinent data for investigating the influence of polymer modifiers on the properties of HMA specimens, particularly rutting resistance and low temperature cracking. Since many mixture properties and strength characteristics have been pointed out in the literature that would relate to the rutting performance, regression analysis was utilized to develop mathematical relationships between rut depth and HMAs physical properties and strength properties. Three parameters obtained from the indirect tensile test, resilient 151

modulus test, and dynamic modulus test are correlated to low temperature cracking properties obtained from the TSRST test. Based on the analysis results, the following conclusions can be made. 1. Adding polymers to virgin binder enhanced the mechanical properties of the HMA including both indirect tensile strength and unconfined compressive strength. Modifier content, by weight of 3 to 4 percent appears to provide best results. 2. According to the APA test results, asphalt concrete mixtures containing modifiers exhibit significantly improved rutting resistance, compared to the unmodified asphalt concrete mixtures. 3. Rut depth decreased steeply with the increase of the polymer content up to 4 percent. Modifier content of greater than 4% does not to seem to offer any additional benefits in enhancing rutting resistance. An optimum polymer content of 3 to 4 percent is observed. 4. From the summary of the results, no strong correlation was found between the physical mixture properties and the rutting performance. Physical mixture properties cannot explain the rutting performance observed from APA tests. 5. Poor correlation (R2-value=0.392) was obtained between the indirect tensile strength and the rut depth. A better correlation (R2-value=0.563) was found between the unconfined compressive strength and rut depth. 6. The study using the stepwise regression method showed that the rut depth can be statistically related to unconfined compressive strength and VMA with R-square of 0.842. 152

7. SBS modifiers perform better than the SBR modifiers in improving the tensile strength in the Thermal Stress Restrained Specimen Test. SBS Koch modified

mixtures have shown the best in improving the fracture temperature. Maximum improvement in fracture temperature and the tensile strength improvement are achieved at 4 percent modifier content. 8. From statistical analysis using regression techniques, only indirect tensile strength has a fair correlation to the measured TSRST fracture strength, the resilient modulus and dynamic modulus did not provide a good definitive relationship with TSRST fracture strength. The parameters obtained from the indirect tensile test, resilient modulus test, and dynamic modulus test have no relationships with TSRST fracture temperature. 9. The parameters obtained from the HMAs simple laboratory test is not enough to evaluate the HMAs low temperature cracking properties, the properties of the asphalt should be considered.

153

REFERENCES

1. American Association of State Highway and Transportation Officials (2001), Standard Specifications for Transportation Materials and Methods of Sampling and Testing, AASHTO. 2. American Society of Testing and Materials Standards (2001), ASTM D 4867/D 4867 M 96. 3. Anani, A. B., (1990),Laboratory and Field Study of Pavement Rutting in Saudi Arabia, Transportation Research Record, 1259, National Research Council, Washington D.C, pp 79-90. 4. Anderson, R.M., (2002), Using Superpave Gyratory Compaction Properties to Estimate the Rutting Potential of Asphalt Mixtures, Asphalt Paving Technology, Association of Asphalt Paving Technologists-Proceedings of the Technical Sessions, Vol.71, pp 725-738. 5. Arand, W. (1987), Influence of bitumen hardness on the fatigue behavior of asphalt pavements of different thickness due to bearing capacity of subbase, traffic loading, and temperature. Proceedings, Sixth International Conference on Structural Behavior of Asphalt Pavements, University of Michigan, pp. 65-71. 6. Bilal A. Abu Alfoul, (2004), Anisotropy and Permanent Deformation of Hot Mix Asphalt A dissertation presented to The University of Akron. 7. Bouzid Choubane, Gale C. Page and James A. Musselman (2000), Effects of Different Water Saturation Levels on the Resistance of Compacted HMA Samples to Moisture Induced Damage Journal of the Transportation Research Board 1723, Washington, D.C., Pp 97-106. 8. Brown, E. R. , and S. A. Cross. (1989), A Study of In-Place Rutting of Asphalt Pavement Association of Asphalt Paving Technologists, Vol. 58, pp.1-31. 9. Buchanan, M., Moore, V., Mallick, R., OBrien, S., and Regimand, A. (2004), Accelerated Moisture Susceptibility Testing of Hot Mix Asphalt (HMA) Mixes, 83rd Transportation Research Board Annual Meeting, Washington, D.C. 154

10. CADOT Transportation Library (2003), Method of Test for Resistance of Compacted Bituminous Mixture to Moisture Induced Damage, California Test 371. 11. Carpenter, S. H. (1983), Thermal cracking in asphalt pavements: an examination of models and input parameters. United States Army Corps of Engineers Cold Regions Research and Engineering Laboratory. 12. Castro-Fernandez, P., Cervantes, V., Barrantes,R., Blackamore, V., and Ingenieria, E. (2004), Effect of Aging on Asphalt Concrete Moisture Damage, 83rd Transportation Research Board Annual Meeting, Washington, D.C. 13. Changlin Pan and Thomas D. White (1998), Evaluation of Stripping for Asphalt Concrete Mixtures Using Accelerated Testing Methods, Transportation Research Record 1630, Washington, D.C., Pp. 98 105. 14. Changlin Pan and Thomas D. White (1999), Conditions for Stripping using Accelerated Testing, FHWA Final Report IN/JTRP-97/13, Purdue University, West Lafayette, Indiana. 15. Choubane, B., Page G., and Musselman, J. (2000), Effects of Different Water Saturations Levels on the Resistance of Compacted HMA Samples to Moisture Induced Damage, Transportation Research Record 1723, Washington, D.C., Pp. 97-106. 16. Christensen, W. D., and Bonaquist, R., (2002), Use of Strength Tests for Evaluating the Rut Resistance of Asphalt Concrete, Asphalt Paving Technology, Association of Asphalt Paving Technologists-Proceedings of the Technical Sessions, Vol.71, pp 692-711. 17. Cominsky, R.J., Huber,G.A., Kennedy, T.W., and Anderson, M. (1994), The Superpave Mix Design Manual for New Construction and Overlays, SHRP-A-407, Strategic Highway Research Program, National Research Council, Washington, D.C.,. 18. Cross, S., Voth, M. and Fager, G. (2000), Effects of Sample Preconditioning on Asphalt Pavement Analyzer Wet Rut Depths, Mid-Continent Transportation Symposium 2000 Proceedings, Transportation Research Board, Washington, D.C. 19. Donald W. Christensen, Ramon Bonaquist, David A. Anderson and Salil Gokhale, (2000), Indirect Tension Strength as a Simple Performance Test, Transportation Research Circular E-C068: New Simple Performance Test for Asphalt Mixes. 20. Elliott, R. P., and Thornton, S. L., (1998), Resilient modulus and AASHTO pavement design Transportation Research Record No. 11969, pp. 116- 124.

155

21. Epps, J., Sebaaly, P., Penaranda, J., Maher, M., McCann, M., and Hand, A. (2000) , Compatibility of a Test for Moisture-Induced Damage with Superpave Volumetric Mix Design, NCHRP Report 444, TRB, Washington, D.C. 22. Fabb, T. R. J. (1974), The influence of mix composition, binder properties and cooling rate on asphalt cracking at low temperature. Proceedings, AAPT, Vol. 43, pp. 285-331. 23. FDOT Specifications (2002), Florida Method of Test for Resistance of Compacted Bituminous Mixture to Moisture-Induced Damage, Designation FM 1-T 283. 24. Feng, A., Hua, J., and White, V, (1999), Flexible Pavement Drainage Monitoring Performance and Stability Purdue University, West Lafayette. 25. Harm, E. and Lippert, D. (2001), Illinois Extended Life HMA Pavement Specifications, Transportation Research Circular 503, Transportation Research Board, Washington, D.C. 26. Hicks, R. (1991), Moisture Damage in Asphalt Concrete, NCHRP Synthesis of Highway Practice, TRB, Washington, D.C. 27. Hicks, R., Leahy, R., Cook, M., Moulthrop, J., and Button, J. (2004), Road Map for Mitigating National Moisture Sensitivity Concerns in Hot Mix Pavement, 83rd Transportation Research Board Annual Meeting, Washington, D.C. 28. Huber, G. A. , and G. H. Heiman. (1987), Effect of Asphalt Concrete Parameters on Rutting Performance: A Field Investigation, Association of Asphalt Paving Technologists, Vol. 56, pp. 36-61. 29. Hunter, R. and Ksaibati, K. (2002), Evaluating Moisture Susceptibility of Asphalt Mixes, University of Wyoming, Laramie, Wyoming. 30. INDOT Specifications (1999), Section 400: Asphalt Pavement, Pp.263-266. 31. John P. Zaniewski, and Geetha Srinivasan, (2004), Evaluation of Indirect Tensile Strength to Identify Asphalt Concrete Rutting Potential, Asphalt Technology Program, Department of Civil and Environmental Engineering, West Virginia University, Morgantown, West Virginia. 32. Jadoun, F. M. (2001), Performance Characteristics and Behavior of PolymerModified Asphalt Concrete Mixture. MSc. Thesis, The University of Akron. 33. Khosla, N., Birdsall, B. and Kawaguchi, S. (2000), Evaluation of Moisture Susceptibility of Asphalt Mixtures Using Conventional and New Methods, Transportation Research Record 1728, Washington, D.C., Pp. 43 51. 156

34. Lottman, R. P. (1978), NCHRP Report 192: Predicting Moisture-Induced Damage to Asphaltic Concrete. Transportation Research Board, National Research Council, Washington, D.C.,. 35. Lottman, R. P. (1982), NCHRP Report 246: Predicting Moisture-Induced Damage to Asphaltic ConcreteField Evaluation. Transportation Research Board, National Research Council, Washington, D.C.,. 36. L. Y. Zhu, T. F. Fwa and Yurong Liu, (2001), Rutting potential Evaluation of Asphalt Mixtures by Repeated-Load Creep Test, 2002 Annual Meeting of Transportation Research Record. 37. McCann, M. and Sebaaly, P. (2001), A Quantitative Evaluation of Stripping Potential in Hot Mix Asphalt Using Ultrasonic Energy for Moisture Accelerated Conditioning, Transportation Research Record 1767, Washington, D.C., Pp. 48 59. 38. McGennis, R., Michale, R., Perdomo, D., and Turner, P. (1996), Issues Pertaining to Use of Superpave Gyratory Compactor, Transportation Research Record 1543, Washington, D.C., Pp. 139-144 39. Monismith, C. L. (1976), Rutting prediction in asphalt concrete pavements Transportation Research Record, 616, pp. 2-8. 40. Monismith, C. L., G. A. Secor, and K. E. Secor. (1965), Temperature induced stresses and deformations in asphalt concrete. Proceedings, AAPT, Vol. 34, pp. 248-285. 41. Musselman, J., Choubane, B., Page, G., and Upshaw, P. (1998), Superpave Field Implementation: Floridas Early Experience, Transportation Research Record 1609, Washington, D.C., Pp. 51 60. 42. M. W. WITCZAK et al. (2002), Simple Performance Test for Superpave Mix Design, NCHRP Report 465. 43. Myron Thiessen, Ahmed shalaby, and Leonnie Kavanagh, (2000), Strength Test of In-Service Asphalt Pavements in Manitoba and Correlation to Rutting, Proceeding of Canadian Technical Asphalt Association 45 pp 203-227. 44. NCDOT TSR Test Procedure (1999), AASHTO T 283 (Modified). 45. NCHRP Active Project 9-34 (2004), Improved Conditioning Procedure for Predicting the Moisture Susceptibility of HMA Pavements, NCHRP-Active Project, TRB, Washington, D.C.

157

46. Pedro Castro-Fernandez, Peter E. Sebaaly, Fabricio Leiva-Villacorta, (2006), Multivariate Analysis on Rutting of Costa Rican Asphalt Concrete Mixes 2006 Annual Meeting of Transportation Research Record. 47. NYDOT Specifications (2002), Superpave Hot Mix Asphalt Design and Mixture Verification Procedures / Materials Method No. NY 5.16. 48. Peter E. Sebaaly, Andrew Lake, and Jon Epps, (2002), Evaluation of LowTemperature Properties of HMA Mixtures Journal of transportation engineering, Vol. 128, pp. 578-586. 49. Tunnicliff, D. G., and R. E. Root. (1984), NCHRP Report 274: Use of Antistripping Additives in Asphaltic Concrete Mixtures Laboratory Phase. Transportation Research Board, National Research Council, Washington, D.C.,. 50. Tunnicliff, D. G., and R. E. Root. (1995), NCHRP Report 373: Use of Antistripping Additives in Asphaltic Concrete MixturesField Evaluation. 51. Rebecca S.McDaniel, Hussain U. Bahia, (2002), Field Evaluation of Asphalt Additives to Control Rutting and Cracking, 2003 Annual Meeting of Transportation Research Record. 52. Roberts, F., Kandhal, P., and Brown, E. (1996), Text Book: Hot Mix Asphalt Materials, Mixture Design, and Construction, Second Edition, NAPA Education Foundation Lanham, Maryland. 53. Scherocman, J. a., Keith, A., M, and Joseph, J. P., (1986), The effect of multiple freeze-thaw cycle conditioning on the moisture damage in asphalt concrete mixtures Asphalt Paving Technology, Vol. 55, pp. 213-227. 54. Stacy G. Williams, et al. (2002), The Effects of HMA Mixture Characteristics on Rutting Susceptibility, 2003 Annual Meeting of Transportation Research Record. 55. Sugawara, T., H. Kubo, and A. Moriyoshi. (1982), Low temperature cracking of asphalt pavements. Proceedings, Workshop in Paving in Cold Areas, Vancouver, B.C., Vol.1, pp. 1-42. 56. Tan, S. A., Low, B. H., and Faw, T.F., (1994), Behavior of asphalt concrete mixtures in triaxial compression Journal of Testing and Evaluation, JTEVA, Vol. 22, No. 3, pp. 169-203. 57. Thom, N. H., and Brown, S. F., (1987),"The effect of moisture on the structural performance of a crushed limestone road base", Annual Transportation Research Board Meeting, Washington, D. C.

158

58. Timothy R. Clyne et al. (2006), Investigation of the Low-Temperature Fracture Properties of Three MnROAD Asphalt Mixtures Report no. Mn/DOT 2006-15. 59. Thompson M. R. and Neumann D. (1993), Rutting rate analyses of the AASHTO road test flexible pavement Transportation Research Record, 1384, pp. 36-48. 60. Ullditz, P. (1993), Mathematical model for pavement performance under moving wheel load Transportation Research Record, 1384, pp. 94-99. 61. UTDOT Materials Manual (2003), Section 957: Resistance of Compacted Bituminous Mixture to Moisture Induced Damage. 62. Uzan, J. (1985), Characterization of granular material Transportation Research Record, 1022, pp. 52-59. 63. Watson (2003), an Updated Review of SMA and Superpave Projects, Paper submitted for presentation and publication at the 2003 Annual Meeting of the Transportation Research Board, Washington D.C. 64. Van de Loo, P. J. (1978), The creep test, a key tool in asphalt mix design and in the prediction of pavement rutting Proceeding Association of Asphalt Paving Technologists, No.47, pp.253-281. 65. Vinson et al. (1990), Summary Report on Low Temperature and Thermal Figure Cracking SHRP-A/IR-90-001, National Research Council, Washington, D. C.

159

APPENDICES

160

APPENDIX A TEST RESULTS OF THE COMPLETE FACTORIAL EXPERIMENTAL PROGRAM A1. Complete Factorial Experimental Program Test Results for Honey Creek Limestone

Table A-1 Test Results for Marshall 100 Specimen with No Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning Sample number 1 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 2 3

deviation samples 2.32 3.35 4.75 0.58 2.32 3.34 4.75 1.75 2.32 0.65 4.75 3.33 3 3 3 3 3 3 3 3 3 3 3 3

87.54 90.83 92.01 90.13 71.09 73.53 66.9 70.51 105.2110.93114.62110.25 79.82 79.04 80.18 79.68 87.54 90.83 92.01 90.13 57.5 64.14 60.16 60.60 105.2110.93114.62110.25 69.62 72.94 72.23 71.60 87.54 90.83 92.01 90.13 51.3 50.23 51.4 50.98 105.2110.93114.62110.25 61.9 68.55 64.81 65.09

50

0.782

0.723

No Aging

70

0.672

0.649

90

0.566

0.590

161

Table A-2 Test Results for Marshall 100 Specimen with 4 Hrs Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 4.38 1.98 5.61 3.67 4.38 3 3 3 3 3 3 3 3 3 3 3 3

161.42167.62169.89166.31 167.6 168 171.21168.94 187.06194.47198.07193.20 166.34170.08162.75166.39 161.42167.62169.89166.31

1.015

0.861

4 hrs Aging

70

129.37 148.2 145.43141.00 10.17 187.06194.47198.07193.20 150.51147.88136.67145.02 161.42167.62169.89166.31 116.97123.13110.74116.95 187.06194.47198.07193.20 120.09116.45114.02116.85 5.61 7.35 4.38 6.20 5.61 3.06

0.847

0.750

0.703

0.604

162

Table A-3 Test Results for Marshall 100 Specimen with 16 Hrs Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 3.70 3 3 3 3 3 3 3 3 3 3 3 3

172.9 166.69173.28170.96

150.93152.75181.08161.59 16.91 170.11 180.3 171.29173.90 156.21150.59 160.6 155.80 172.9 166.69173.28170.96 133.52149.56133.75138.94 170.11 180.3 171.29173.90 123.89124.98137.65128.84 172.9 166.69173.28170.96 115.27117.62128.29120.39 170.11 180.3 171.29173.90 5.57 5.02 3.70 9.20 5.57 7.65 3.70 6.94 5.57

0.945

0.896

16 hrs Aging

70

0.813

0.741

0.704

108.54 135.1 97.26 113.63 19.43

0.653

163

Table A-4 Test Results for Gyratory 100 Specimen with No Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning Sample number 1 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 2 3

deviation samples 3.04 2.44 4.55 0.77 3.04 3.38 4.55 3.81 3.04 1.40 4.55 2.29 3 3 3 3 3 3 3 3 3 3 3 3

78.5780.4084.50 81.16 73.1070.2075.04 72.78 80.9088.8180.97 83.56 75.8774.9074.35 75.04 78.5780.4084.50 81.16 67.9073.1666.86 69.31 80.9088.8180.97 83.56 58.6266.1963.12 62.64 78.5780.4084.50 81.16 57.4059.8859.77 59.02 80.9088.8180.97 83.56 49.8652.8554.37 52.36

50

0.897

0.898

No Aging

70

0.854

0.750

90

0.727

0.627

164

Table A-5 Test Results for Gyratory 100 Specimen with 4 Hrs Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 8.06 5.75 7.45 3.85 8.06 4.98 7.45 6.96 8.06 1.30 7.45 3.93 3 3 3 3 3 3 3 3 3 3 3 3

156.40169.56171.05165.67 144.22147.22155.33148.92 172.88166.75181.58173.74 152.04159.64156.87156.18 156.40169.56171.05165.67 133.22123.69130.95129.29 172.88166.75181.58173.74 132.24141.95145.74139.98 156.40169.56171.05165.67 110.56110.47112.76111.26 172.88166.75181.58173.74 124.60118.49125.82122.97

0.899

0.899

4 hrs Aging

70

0.780

0.806

0.672

0.708

165

Table A-6 Test Results for Gyratory 100 Specimen with 16 Hrs Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 3.63 3.03 5.28 5.38 3.63 3 3 3 3 3 3 3 3 3 3 3 3

162.32162.83 156.3 160.48 135.11 131.9 137.95134.99 165.35174.37 165.1 168.27 150.96144.85155.58150.46 162.32162.83 156.3 160.48

0.841

0.894

16 hrs Aging

70

114.87136.15128.05126.36 10.74 165.35174.37 165.1 168.27 136.19138.16134.52136.29 162.32162.83 156.3 160.48 115.8 105.46103.98108.41 165.35174.37 165.1 168.27 126.54109.13111.53115.73 5.28 1.82 3.63 6.44 5.28 9.44

0.787

0.810

0.676

0.688

166

Table A-7 Test Results for Gyratory 150 Specimen with No Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning Sample number 1 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 2 3

deviation samples 3.57 1.45 2.79 1.89 3.57 5.45 2.79 6.54 3.57 3.92 2.79 3.36 3 3 3 3 3 3 3 3 3 3 3 3

78.8284.9185.10 82.94 70.3469.6267.54 69.17 83.9384.6979.52 82.71 72.0769.3768.44 69.96 78.8284.9185.10 82.94 63.4755.7552.94 57.39 83.9384.6979.52 82.71 59.9455.4947.06 54.16 78.8284.9185.10 82.94 55.8348.5849.64 51.35 83.9384.6979.52 82.71 46.9040.5145.48 44.30

50

0.834

0.846

No Aging

70

0.692

0.655

90

0.619

0.536

167

Table A-8 Test Results for Gyratory 150 Specimen with 4 Hrs Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 7.50 4.48 5.93 5.32 7.50 0.68 5.93 4.98 7.50 4.60 5.93 4.34 3 3 3 3 3 3 3 3 3 3 3 3

121.18135.88125.96127.67 118.59122.48127.52122.86 137.60149.46143.53143.53 136.72 126.2 132.85131.92 121.18135.88125.96127.67 112.33112.04111.03111.80 137.60149.46143.53143.53 112.76121.76113.57116.03 121.18135.88125.96127.67 99.13 94.94 89.95 94.67 137.60149.46143.53143.53 109.45114.89106.32110.22

0.962

0.919

4 hrs Aging

70

0.876

0.808

0.742

0.768

168

Table A-9 Test Results for Gyratory 150 Specimen with 16 Hrs Loose Mix Aging (Honey Creek Limestone)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 9.87 2.38 9.06 2.19 9.87 3.74 9.06 6.08 9.87 1.52 9.06 3.10 3 3 3 3 3 3 3 3 3 3 3 3

146.15130.37148.54141.69 141.75142.95138.36141.02 161.22177.77 175.9 171.63 150.67154.99 153.5 153.05 146.15130.37148.54141.69 121.96115.24115.75117.65 161.22177.77 175.9 171.63 125.7 137.68129.87131.08 146.15130.37148.54141.69 102.2 101.78104.59102.86 161.22177.77 175.9 171.63 120.28122.24116.16119.56

0.995

0.892

16 hrs Aging

70

0.830

0.764

0.726

0.697

169

A2. Complete Factorial Experimental Program Test Results for Martin Marietta Gravel

Table A-10 Test Results for Marshall 100 Specimen with No Loose Mix Aging (Martin Marietta Gravel)

Loose Saturation Compacted mix level aging Dry F-T 72 h Dry F-T 24 h No Aging 70 72 h Dry F-T Dry F-T 24 h 90 72 h Dry F-T Dry F-T mix aging conditioning Sample

Tensile strength Statistics (psi) Sample number 1 24 h 50 2 3 Mean Standard N

Tensile strength ratio

deviation samples 2.78 0.68 2.38 1.12 2.78 4.39 5.39 9.93 2.78 1.18 5.39 1.39 3 3 3 3 3 3 3 3 3 3 3 3 0.76 0.82 0.86 0.82 0.97 0.95

83.9289.2085.0186.04 81.2981.6882.6281.86 89.2092.3093.8791.79 89.5387.3788.9888.63 83.9289.2085.0186.04 65.5372.6073.5870.57 89.2092.3081.8187.77 86.4567.5371.7775.25 83.9289.2085.0186.04 70.9170.8768.8570.21 89.2092.3081.8187.77 67.6266.5964.8766.36

170

Table A-11 Test Results for Marshall 100 Specimen with 4 Hrs Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 8.18 3 3 3 3 3 3 3 3 3 3 3 3

180.77164.51171.05172.11

123.18129.68148.52133.80 13.16 191.22175.39195.27187.29 10.51 156.83137.66156.45150.31 10.96 180.77164.51171.05172.11 134.61146.18147.43142.74 8.18 7.07

0.777

0.803

4 hrs Aging

70

0.829

191.22175.39195.27187.29 10.51 158.32157.57166.38160.76 180.77164.51171.05172.11 157.04165.42159.72160.73 4.88 8.18 4.28

0.858

0.934

191.22175.39195.27187.29 10.51 157.04164.68157.77159.83 4.22

0.853

171

Table A-12 Test Results for Marshall 100 Specimen with 16 Hrs Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 4.05 5.33 7.43 3 3 3 3 3 3 3 3 3 3 3 3

138.04129.96133.47133.83 171.95162.02163.60165.86 150.77160.93165.25158.99

1.239

138.53136.92157.10144.18 11.21 138.04129.96133.47133.83 4.05

0.907

16 hrs Aging

70

184.18159.84154.37166.13 15.87 150.77160.93165.25158.99 165.20158.21150.24157.88 138.04129.96133.47133.83 151.51146.24142.22146.66 150.77160.93165.25158.99 153.37145.06142.38146.94 7.43 7.48 4.05 4.66 7.43 5.73

1.241

0.993

1.096

0.924

172

Table A-13 Test Results for Gyratory 100 Specimen with No Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning Sample number 1 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 2 3

deviation samples 0.91 3.13 0.44 1.11 0.91 1.40 0.44 4.23 0.91 3.79 0.44 1.47 3 3 3 3 3 3 3 3 3 3 3 3

80.0078.5978.32 78.97 75.0969.2670.20 71.51 78.7977.9378.49 78.40 68.9868.5266.86 68.12 80.0078.5978.32 78.97 67.6467.1469.77 68.18 78.7977.9378.49 78.40 59.9967.3767.25 64.87 80.0078.5978.32 78.97 67.7560.1863.71 63.88 78.7977.9378.49 78.40 71.6671.1268.90 70.56

50

0.906

0.869

No Aging

70

0.863

0.827

90

0.809

0.900

173

Table A-14 Test Results for Gyratory 100 Specimen with 4 Hrs Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 2.83 7.69 1.65 5.98 2.83 5.40 1.65 3 3 3 3 3 3 3 3 3 3 3 3

148.30153.83152.11151.41 145.22130.30140.99138.84 145.07142.78146.00144.62 132.92123.58134.72130.41 148.30153.83152.11151.41 144.33135.90145.94142.06 145.07142.78146.00144.62

0.917

0.902

4 hrs Aging

70

0.938

127.26132.57162.48140.77 18.99 148.30153.83152.11151.41 2.83

0.973

155.79135.06147.16146.00 10.42 145.07142.78146.00144.62 1.65

0.964

156.66129.31157.20147.72 15.95

1.021

174

Table A-15 Test Results for Gyratory 100 Specimen with 16 Hrs Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 3 3 3 3 3 3 3 3 3 3 3 3

127.57146.06146.69140.10 10.86 151.70157.53155.54154.92 2.96

1.106

137.93143.18161.34147.48 12.28 159.75153.56147.89153.73 5.93

1.042

127.57146.06146.69140.10 10.86 129.37141.81114.53128.57 13.66 137.93143.18161.34147.48 12.28 133.69132.58148.40138.22 8.83

16 hrs Aging

70

0.918

0.937

127.57146.06146.69140.10 10.86 133.10112.86132.01125.99 11.38 137.93143.18161.34147.48 12.28 161.91135.91144.32147.38 13.26

0.899

0.999

175

Table A-16 Test Results for Gyratory 150 Specimen with No Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning Sample number 1 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h Dry F-T 72 h Dry F-T 2 3

deviation samples 7.30 3.06 1.52 3.81 7.30 5.07 1.52 3.24 7.30 3.25 1.52 3.67 3 3 3 3 3 3 3 3 3 3 3 3

59.9664.8050.45 58.40 64.3761.7658.28 61.47 70.2871.1973.25 71.57 61.4566.5568.90 65.63 59.9664.8050.45 58.40 34.7144.8539.78 39.78 70.2871.1973.25 71.57 52.9058.7658.21 56.62 59.9664.8050.45 58.40 42.6444.9149.04 45.53 70.2871.1973.25 71.57 54.8747.5550.93 51.12

50

1.053

0.917

No Aging

70

0.681

0.791

90

0.780

0.714

176

Table A-17 Test Results for Gyratory 150 Specimen with 4 Hrs Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 6.85 3 3 3 3 3 3 3 3 3 3 3 3

168.60160.74174.39167.91

139.21158.25168.20155.22 14.73 153.18142.73144.98146.96 112.70128.71122.37121.26 168.60160.74174.39167.91 167.58161.02175.97168.19 153.18142.73144.98146.96 121.50127.10122.62123.74 168.60160.74174.39167.91 153.33171.58161.19162.03 153.18142.73144.98146.96 112.66104.80121.59113.01 5.50 8.07 6.85 7.50 5.50 2.96 6.85 9.15 5.50 8.40

0.924

0.825

4 hrs Aging

70

1.002

0.842

0.965

0.769

177

Table A-18 Test Results for Gyratory 150 Specimen with 16 Hrs Loose Mix Aging (Martin Marietta Gravel)

Loose mix aging

Tensile strength Saturation Compacted level Sample (psi) Sample number 1 24 h 50 72 h Dry F-T Dry F-T 24 h Dry F-T 72 h Dry F-T 24 h 90 72 h Dry F-T Dry F-T 2 3 Mean

Statistics Standard N

Tensile strength ratio

mix aging conditioning

deviation samples 7.39 3 3 3 3 3 3 3 3 3 3 3 3

133.11124.14138.80132.02

109.28129.76120.40119.81 10.25 149.89152.77150.07150.91 147.84145.31140.78144.64 133.11124.14138.80132.02 161.40154.03170.87162.10 149.89152.77150.07150.91 1.61 3.58 7.39 8.44 1.61

0.908

0.958

16 hrs Aging

70

1.228

112.35128.93139.07126.78 13.49 133.11124.14138.80132.02 130.87140.06142.37137.77 149.89152.77150.07150.91 7.39 6.08 1.61

0.840

1.044

104.29132.30144.76127.12 20.73

0.842

178

APPENDIX B OPTIMUM ASPHALT CONTENT MIX DESIGN

1. Honey Creek Limestone

Marshall mix design Gyratory 4 inch mix design Gyratory 6 inch mix design

2. Martin Marietta Gravel

Marshall mix design Gyratory 4 inch mix design Gyratory 6 inch mix design

3. Traprock Source
Marshall mix design Gyratory 6 inch mix design

4. Stocker Sand Gravel

Marshall mix design Gyratory 6 inch mix design

179

B1. Honey Creek Limestone

Marshall mix design

Table B-1 Marshall mix design for Honey Creek Limestone

Honey Creek Limestone Unit Weight, % AC by wt. of mix. Pcf, (Mg/m3) 4.5 5 5.5 6 6.5 7 149.53 149.64 150.54 150.67 150.87 151.19 4.91 4.19 3.46 2.72 2.47 2.14 13.64 14.03 13.97 14.35 14.69 14.97 64.02 70.12 75.24 81.06 83.20 85.68 % Air Voids % VMA% VFA lbs(N) 2274 2250 2403 2057 2253 1962 11.3 11.6 12.1 12.5 13.3 13.7 Stability, Flow 0.01 in. (0.25 mm)

Figure B-1 Marshall Mix Design (AV Vs. AC for Honey Creek Limestone)
Air Voids, %
6.0 5.0 Optim um AC=5.1% @ 4.0 % Air voids

Air Voids, %

4.0 3.0 2.0 1.0 0.0 3.5 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

180

Figure B-2 Marshall Mix Design (VMA Vs. AC for Honey Creek Limestone)
VMA, % 15.2 15 14.8 VMA, % 14.6 14.4 14.2 14 13.8 13.6 13.4 3.5 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

Figure B-3 Marshall Mix Design (Stability Vs. AC for Honey Creek Limestone)
S tability, pounds 2450 2400 2350 2300 2250 2200 2150 2100 2050 2000 3.5 4 4.5 5 5.5 6 6.5 7 7.5

Stability, Pounds

Asphalt Content, %

181

Figure B-4 Marshall Mix Design (Flow Vs. AC for Honey Creek Limestone)
Flow, 0.01 in 16 14 Flow, 0.01 in 12 10 8 3.5 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

Figure B-5 Marshall Mix Design (VFA Vs. AC for Honey Creek Limestone)
VFA, % 90.0 85.0 80.0 VFA, % 75.0 70.0 65.0 60.0 55.0 50.0 3.5 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

182

Figure B-6 Marshall Mix Design (Unit WGT Vs. AC for Honey Creek Limestone)
Unit Weight, pcf 151.4 151.2 151 Unit weight, pcf 150.8 150.6 150.4 150.2 150 149.8 149.6 149.4 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

183

 Gyratory 4 inch optimum asphalt mix design

Table B-2 Compaction Data for 5% AC (G4 for Honey Creek Limestone)
Densification Data @ 5.0 % Asphalt Content Gmm(means) = 2.492 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(means)= CF= 71.6 70.6 69.5 67.4 66.2 65.5 64.9 64.4 64.2 64 63.7 63.5 63.1 562.345 554.491 545.851 529.358 519.933 514.435 509.723 505.796 504.225 502.654 500.298 498.727 495.586 1165.3 2.414 1.027 2.072 2.102 2.135 2.201 2.241 2.265 2.286 2.304 2.311 2.318 2.329 2.337 2.351 2.128 2.158 2.192 2.260 2.301 2.326 2.347 2.365 2.373 2.380 2.391 2.399 2.414 85.4 86.6 88.0 90.7 92.3 93.3 94.2 94.9 95.2 95.5 96.0 96.3 96.9 72.1 71.2 70.1 68.2 67.1 66.3 65.7 65.3 65.1 64.9 64.6 64.3 64 566.272 559.203 550.564 535.641 527.002 520.719 516.006 512.865 511.294 509.723 507.367 505.011 502.654 1194.3 2.438 1.026 2.109 2.136 2.169 2.230 2.266 2.294 2.315 2.329 2.336 2.343 2.354 2.365 2.376 2.164 2.192 2.226 2.288 2.326 2.354 2.375 2.390 2.397 2.405 2.416 2.427 2.438 86.9 86.114 88.0 87.268 89.3 88.643 91.8 91.258 93.3 92.832 94.5 93.889 95.3 94.752 95.9 95.409 96.2 95.704 96.5 96.001 96.9 96.450 97.4 96.828 97.8 97.361 Specimen 2 Avg

184

Table B-3 Compaction Data for 5.5% AC (G4 for Honey Creek Limestone)
Densification Data @ 5.5 % Asphalt Content Gmm(means) = 2.491 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(means)= CF= 71.5 70.3 69.4 67.4 66.3 65.5 64.9 64.5 64.3 64.1 63.8 63.5 63.2 561.559 552.134 545.066 529.358 520.719 514.435 509.723 506.581 505.011 503.440 501.084 498.727 496.371 1181.5 2.439 1.025 2.104 2.140 2.168 2.232 2.269 2.297 2.318 2.332 2.340 2.347 2.358 2.369 2.380 2.156 2.192 2.221 2.287 2.325 2.353 2.375 2.389 2.397 2.404 2.416 2.427 2.439 86.532 70.5 88.009 69.6 89.151 68.5 91.796 66.5 93.319 65.4 94.459 64.7 95.332 64.1 95.923 63.7 96.222 63.5 96.522 63.3 96.976 63 553.705 546.637 537.997 522.289 513.650 508.152 503.440 500.298 498.727 497.157 494.800 493.230 490.088 1165.3 2.438 1.025 2.105 2.132 2.166 2.231 2.269 2.293 2.315 2.329 2.337 2.344 2.355 2.363 2.378 2.158 2.186 2.221 2.288 2.326 2.351 2.373 2.388 2.396 2.403 2.415 2.422 2.438 86.623 86.577 87.743 87.876 89.152 89.151 91.833 91.815 93.378 93.348 94.388 94.423 95.271 95.302 95.870 95.896 96.172 96.197 96.476 96.499 96.935 96.955 97.244 97.339 97.867 97.882 Specimen 2 Avg

97.434 62.8 97.896 62.4

185

Table B-4 Compaction Data for 6% AC (G4 for Honey Creek Limestone)
Densification Data @ 6.0 % Asphalt Content Gmm(means) = 2.468 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr) %GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(means)= CF= 71.2 70.2 69.1 67 65.9 65.1 64.5 64 63.8 63.7 63.4 63.1 62.7 559.203 551.349 542.710 526.216 517.577 511.294 506.581 502.654 501.084 500.298 497.942 495.586 492.444 1166.6 2.431 1.026 2.086 2.116 2.150 2.217 2.254 2.282 2.303 2.321 2.328 2.332 2.343 2.354 2.369 2.141 2.172 2.206 2.275 2.313 2.342 2.364 2.382 2.390 2.393 2.405 2.416 2.431 86.757 71.6 87.993 70.6 89.394 69.5 92.195 67.4 93.734 66.2 94.886 65.4 95.769 64.8 96.517 64.4 96.820 64.2 96.972 64 562.345 554.491 545.851 529.358 519.933 513.650 508.938 505.796 504.225 502.654 500.298 497.942 495.586 1173.3 2.440 1.031 2.086 2.116 2.149 2.216 2.257 2.284 2.305 2.320 2.327 2.334 2.345 2.356 2.368 2.150 2.181 2.215 2.284 2.326 2.354 2.376 2.391 2.398 2.405 2.417 2.428 2.440 87.122 86.939 88.356 88.174 89.754 89.574 92.550 92.373 94.228 93.981 95.381 95.133 96.264 96.016 96.862 96.689 97.164 96.992 97.467 97.219 97.926 97.678 98.390 98.142 98.857 98.688 Specimen 2 Avg

97.430 63.7 97.894 63.4 98.518 63.1

186

Table B-5 Compaction Data for 6.5% AC (G4 for Honey Creek Limestone)
Densification Data @ 6.5 % Asphalt Content Gmm(means) = 2.461 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(means)= CF= 72.1 70.8 69.8 67.8 66.6 65.8 65.2 64.8 64.6 64.4 64.1 63.9 63.6 566.272 556.061 548.207 532.500 523.075 516.792 512.079 508.938 507.367 505.796 503.440 501.869 499.513 1198.5 2.444 1.019 2.116 2.155 2.186 2.251 2.291 2.319 2.340 2.355 2.362 2.370 2.381 2.388 2.399 2.156 2.195 2.227 2.293 2.334 2.362 2.384 2.399 2.406 2.414 2.425 2.432 2.444 87.599 70.7 89.207 69.7 90.485 68.6 93.154 66.4 94.833 65.3 95.986 64.4 96.869 63.9 97.467 63.4 97.769 63.2 98.072 63 555.276 547.422 538.783 521.504 512.865 505.796 501.869 497.942 496.371 494.800 492.444 490.873 487.732 1171.6 2.449 1.020 2.110 2.140 2.175 2.247 2.284 2.316 2.334 2.353 2.360 2.368 2.379 2.387 2.402 2.151 2.182 2.217 2.290 2.329 2.362 2.380 2.399 2.406 2.414 2.426 2.433 2.449 87.407 87.503 88.662 88.934 90.083 90.284 93.068 93.111 94.636 94.734 95.958 95.972 96.709 96.789 97.472 97.469 97.780 97.775 98.091 98.082 98.560 98.546 98.875 98.858 99.512 99.409 Specimen 2 Avg

98.531 62.7 98.840 62.5 99.306 62.1

187

Table B-6 Summary of Compaction Data (G4 for Honey Creek Limestone)

Number of Gyration % Gmm, 5.0 AC 5 7 10 20 30 40 50 60 65 70 80 90 99 86.114 87.268 88.643 91.258 92.832 93.889 94.752 95.409 95.704 96.001 96.450 96.828 97.361

% Gmm, 5.5 AC 86.577 87.876 89.151 91.815 93.348 94.423 95.302 95.896 96.197 96.499 96.955 97.339 97.882

%Gmm,6.0AC %Gmm, 6.5 AC 86.939 88.174 89.574 92.373 93.981 95.133 96.016 96.689 96.992 97.219 97.678 98.142 98.688 87.503 88.934 90.284 93.111 94.734 95.972 96.789 97.469 97.775 98.082 98.546 98.858 99.409

Table B-7 Summary of Design Parameters (G4 for Honey Creek Limestone)
% Gmm AC % 5 5.5 6 6.5 @N=7 87.268 87.876 88.174 88.934 % Gmm @N=65 95.704 96.197 96.992 97.775 % Gmm @N=99 97.361 97.882 98.688 99.409 Air Void, %, @ Ndes 4.296 3.803 3.008 2.225 Gsb 2.65 2.65 2.65 2.65 Ps 0.95 0.945 0.94 0.935 Gmm % VMA % VFA 2.492 14.497 2.491 14.537 2.468 15.094 2.461 15.100 70.369 73.836 80.069 85.262

188

Figure B-7 Gyratory 4 Mix Design (Gmm% Vs. Gyration No. for Honey Creek Limestone)
Average Densification Curves
101 100 99 98 97 96 95
5.0 % AC 5.5 % AC 6.0 % AC 6.5 % AC

% Gmm

94 93 92 91 90 89 88 87 86 85 1 10 100 1000

Number of Gyrations

Figure B-8 Gyratory 4 Mix Design (Gmm% at Initial Vs. AC for Honey Creek Limestone)
89.2 89.0

Percent of Gmm at Nini

88.8 88.6 88.4 88.2 88.0 87.8 87.6 87.4 87.2 87.0 4.5 5 5.5 6 6.5 7

Asphalt Content, %

189

Figure B-9 Gyratory 4 Mix Design (Gmm% at Max Vs. AC for Honey Creek Limestone)

100.0 99.5

Percent of Gmm at Nmax

99.0 98.5 98.0 97.5 97.0 96.5 96.0 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-10 Gyratory 4 Mix Design (Air Void Vs. AC for Honey Creek Limestone)
5.0 4.5 4.0 3.5

Air Voids, %

3.0 2.5 2.0 1.5 1.0 0.5 0.0 4.5 5 5.5 6 6.5 7

Asphalt Content, %

190

Figure B-11 Gyratory 4 Mix Design (VMA Vs. AC for Honey Creek Limestone)
16.5 16.0 15.5

VMA, %

15.0 14.5 14.0 13.5 13.0 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-12 Gyratory 4 Mix Design (VFA Vs. AC for Honey Creek Limestone)
90 85 80

VFA, %

75 70 65 60 55 50 4.5 5 5.5 6 6.5 7

Asphalt Content, %

191

 Gyratory 6 inch specimen mix design

Table B-8 Compaction Data for 4% AC (G6 for Honey Creek Limestone)
Densification Data @ 4.0 % Asphalt Content Gmm(means) = 2.561 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(means)= CF= 130.4 128.6 126.7 123.1 121 119.6 118.6 117.8 117.4 117.1 116.6 116.1 115.4 2304.356 2272.548 2238.972 2175.355 2138.245 2113.505 2095.833 2081.696 2074.627 2069.326 2060.490 2051.655 2039.285 4870.5 2.433 1.019 2.114 2.143 2.175 2.239 2.278 2.304 2.324 2.340 2.348 2.354 2.364 2.374 2.388 2.153 2.183 2.216 2.281 2.321 2.348 2.368 2.384 2.392 2.398 2.408 2.418 2.433 84.1 85.3 86.5 89.1 90.6 91.7 92.4 93.1 93.4 93.6 94.0 94.4 95.0 129.9 128.2 126.4 122.9 121 119.6 118.6 117.8 117.4 117.1 116.6 116.1 115.4 2295.521 2265.479 2233.670 2171.820 2138.245 2113.505 2095.833 2081.696 2074.627 2069.326 2060.490 2051.655 2039.285 4829.3 2.433 1.027 2.104 2.132 2.162 2.224 2.259 2.285 2.304 2.320 2.328 2.334 2.344 2.354 2.368 2.162 2.190 2.221 2.285 2.321 2.348 2.368 2.384 2.392 2.398 2.408 2.418 2.433 84.4 84.240 85.5 85.388 86.7 86.636 89.2 89.136 90.6 90.609 91.7 91.670 92.4 92.443 93.1 93.071 93.4 93.388 93.6 93.627 94.0 94.029 94.4 94.434 95.0 95.006 Specimen 2 Avg

192

Table B-9 Compaction Data for 4.5% AC (G6 for Honey Creek Limestone)
Densification Data @ 4.5 % Asphalt Content Gmm(means) = 2.524 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 129.1 127.3 125.5 121.8 119.8 118.5 117.4 116.6 116.3 115.9 115.5 114.9 114.2 2281.383 2249.575 2217.766 2152.382 2117.039 2094.066 2074.627 2060.490 2055.189 2048.120 2041.052 2030.449 2018.079 4893.1 2.459 1.015 2.142 2.172 2.204 2.271 2.309 2.334 2.356 2.372 2.378 2.386 2.394 2.407 2.422 2.175 2.206 2.237 2.305 2.344 2.369 2.392 2.408 2.414 2.422 2.431 2.444 2.459 86.176 129.4 87.394 127.5 88.648 125.7 91.341 121.7 92.866 119.8 93.884 118.5 94.764 117.4 95.414 116.6 95.660 116.3 95.990 115.9 96.323 115.5 96.826 114.9 97.419 114.2 2286.685 2253.109 2221.300 2150.615 2117.039 2094.066 2074.627 2060.490 2055.189 2048.120 2041.052 2030.449 2018.079 4890.5 2.459 1.015 2.138 2.170 2.201 2.273 2.309 2.334 2.356 2.372 2.379 2.387 2.395 2.408 2.422 2.170 2.203 2.234 2.308 2.344 2.370 2.392 2.409 2.415 2.423 2.432 2.444 2.459 85.999 86.087 87.280 87.337 88.530 88.589 91.440 91.390 92.890 92.878 93.909 93.897 94.789 94.777 95.440 95.427 95.686 95.673 96.016 96.003 96.348 96.336 96.852 96.839 97.445 97.432 Specimen 2 Avg

193

Table B-10 Compaction Data for 5% AC (G6 for Honey Creek Limestone)
Densification Data @ 5.0 % Asphalt Content Gmm(meas) = 2.516 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 129.3 127.6 125.7 121.9 119.8 118.4 117.4 116.5 116.1 115.8 115.3 114.9 114.2 2284.918 2254.876 2221.300 2154.149 2117.039 2092.299 2074.627 2058.723 2051.655 2046.353 2037.517 2030.449 2018.079 4894.5 2.461 1.021 2.129 2.158 2.190 2.259 2.298 2.325 2.345 2.363 2.372 2.378 2.388 2.396 2.411 2.173 2.202 2.236 2.305 2.346 2.373 2.394 2.412 2.420 2.427 2.437 2.446 2.461 86.394 129.3 87.545 127.6 88.868 125.7 91.638 121.9 93.244 119.8 94.347 118.4 95.151 117.3 95.886 116.5 96.216 116.1 96.465 115.8 96.884 115.2 97.221 114.8 97.817 114.1 2284.918 2254.876 2221.300 2154.149 2117.039 2092.299 2072.860 2058.723 2051.655 2046.353 2035.750 2028.682 2016.312 4877.1 2.464 1.021 2.130 2.158 2.191 2.259 2.299 2.326 2.348 2.364 2.372 2.378 2.391 2.399 2.414 2.175 2.204 2.237 2.307 2.347 2.375 2.397 2.414 2.422 2.428 2.441 2.449 2.464 86.451 86.422 87.603 87.574 88.927 88.897 91.699 91.669 93.307 93.276 94.410 94.378 95.295 95.223 95.950 95.918 96.280 96.248 96.530 96.498 97.032 96.958 97.371 97.296 97.968 97.892 Specimen 2 Avg

194

Table B-11 Compaction Data for 5.5% AC (G6 for Honey Creek Limestone)
Densification Data @ 5.5 % Asphalt Content Gmm(meas) = 2.493 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 128.6 126.5 124.9 121.1 119 117.6 116.7 115.9 115.6 115.3 114.9 114.6 114.2 2272.548 2235.438 2207.163 2140.012 2102.902 2078.162 2062.257 2048.120 2042.819 2037.517 2030.449 2025.147 2018.079 4931.8 2.482 1.015 2.170 2.206 2.234 2.305 2.345 2.373 2.391 2.408 2.414 2.420 2.429 2.435 2.444 2.204 2.240 2.269 2.340 2.382 2.410 2.428 2.445 2.452 2.458 2.467 2.473 2.482 87.604 123.2 89.059 121.4 90.200 120 2177.122 2145.313 2120.573 2058.723 2023.380 1998.640 1980.969 1966.832 1959.763 1952.695 1943.859 1936.790 1924.420 4651.5 2.464 1.019 2.137 2.168 2.194 2.259 2.299 2.327 2.348 2.365 2.374 2.382 2.393 2.402 2.417 2.178 2.210 2.236 2.303 2.344 2.373 2.394 2.411 2.420 2.429 2.440 2.448 2.464 86.588 87.096 87.871 88.465 88.897 89.548 91.567 92.299 93.167 93.919 94.320 95.059 95.161 95.849 95.845 96.525 96.191 96.824 96.539 97.124 96.978 97.514 97.332 97.819 97.958 98.304 Specimen 2 Avg

93.030 116.5 94.672 114.5 95.799 113.1 96.537 112.1 97.204 111.3 97.456 110.9 97.710 110.5 98.050 110

98.306 109.6 98.651 108.9

195

Table B-12 Summary of Compaction Data (G6 for Honey Creek Limestone)

Number of Gyration 5 7 10 20 30 40 50 60 65 70 80 90 99

% Gmm, 4.0 AC 84.240 85.388 86.636 89.136 90.609 91.670 92.443 93.071 93.388 93.627 94.029 94.434 95.006

% Gmm, 4.5 AC 86.087 87.337 88.589 91.390 92.878 93.897 94.777 95.427 95.673 96.003 96.336 96.839 97.432

% Gmm, 5.0 AC 86.422 87.574 88.897 91.669 93.276 94.378 95.223 95.918 96.248 96.498 96.958 97.296 97.892

% Gmm, 5.5 AC 87.096 88.465 89.548 92.299 93.919 95.059 95.849 96.525 96.824 97.124 97.514 97.819 98.304

Table B-13 Summary of Design Parameters (G6 for Honey Creek Limestone)
% Gmm AC % 4 4.5 5 5.5 @N=7 85.388 87.337 87.574 88.465 % Gmm @N=65 93.388 95.673 96.248 96.824 % Gmm Air Void, Gsb 2.65 2.65 2.65 2.65 Ps 0.96 0.955 0.95 0.945 Gmm % VMA % VFA 2.561 2.523 2.515 2.493 13.359 13.011 13.222 13.923 50.503 66.743 71.624 77.185

@N=99 %, @ Ndes 95.006 97.432 97.892 98.304 6.612 4.327 3.752 3.176

196

Figure B-13 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Honey Creek Limestone)
Average Densification Curves
99 98 97 96 95 94

% Gmm

93 92 91 90 89 88 87 86 85 1 10 100 1000

4.0 % AC 4.5% AC 5.0% AC 5.5% AC

Number of Gyrations

Figure B-14 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Honey Creek Limestone)
89

Percent of Gmm at Nini

89 88 88 87 87 86 86 85 3.5 4 4.5 5 5.5 6

Asphalt Content, %

197

Figure B-15 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Honey Creek Limestone)
98.5

Percent of Gmm at Nmax

98.0 97.5 97.0 96.5 96.0 95.5 95.0 94.5 3.5 4 4.5 5 5.5 6

Asphalt Content, %

Figure B-16 Gyratory 6 Mix Design (Air Void Vs. AC for Honey Creek Limestone)
7

Air Voids, %

0 3.5 4 4.5 5 5.5 6

Asphalt Content, %

198

Figure B-17 Gyratory 6 Mix Design (VMA Vs. AC for Honey Creek Limestone)
14.50 14.25 14.00

VMA, %

13.75 13.50 13.25 13.00 12.75 12.50 3.5 4 4.5 5 5.5 6

Asphalt Content, %

Figure B-18 Gyratory 6 Mix Design (VFA Vs. AC for Honey Creek Limestone)
80

75

70

VFA, %

65

60

55

50 3.5 4 4.5 5 5.5

Asphalt Content, %

199

B2. Martin Marietta Gravel

Marshall specimen mix design

Table B-14 Marshall mix design for Martin Marietta Gravel

Martin Marietta Co % AC by wt. Unit Weight, % Air Voids of mix. 4 4.5 5 5.5 6 6.5 pcf (Mg/m3) 140.88 142.30 143.06 144.21 144.70 143.65 8.15 6.46 5.30 3.55 2.39 2.39 15.86 15.45 15.45 15.22 14.93 15.55 48.62 58.20 65.68 76.69 84.00 84.65 1427 1389 1405 1382 1491 1147 6.3 6.1 6.8 8.7 9.1 10.7 % VMA % VFA Stability (lb) Flow 0.01 in.

200

Figure B-19 Marshall Mix Design (AV Vs. AC for Martin Marietta Gravel)
Air Voids, %
9.0 8.0 7.0 Optim um AC=5.4% @ 4.0 % Air voids

Air Voids, %

6.0 5.0 4.0 3.0 2.0 1.0 0.0 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-20 Marshall Mix Design (VMA Vs. AC for Martin Marietta Gravel)
VMA, %
16.0 15.8 15.6

VMA, %

15.4 15.2 15.0 14.8 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

201

Figure B-21 Marshall Mix Design (Stability Vs. AC for Martin Marietta Gravel)
S tability, pounds
1650 1600 1550 1500 1450 1400 1350 1300 1250 1200 1150 1100 3.5 4 4.5 5 5.5 6 6.5 7

Stability, Pounds

Asphalt Content, %

Figure B-22 Marshall Mix Design (Flow Vs. AC for Martin Marietta Gravel)
Flow, 0.01 in
12 11

Flow, 0.01 in

10 9 8 7 6 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

202

Figure B-23 Marshall Mix Design (VFA Vs. AC for Martin Marietta Gravel)
VFA, %
90 85 80

VFA, %

75 70 65 60 55 50 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-24 Marshall Mix Design (VMA Vs. AC for Martin Marietta Gravel)
Unit Weight, pcf
145 144.5 144

Unit weight, pcf

143.5 143 142.5 142 141.5 141 140.5 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

203

 Gyratory 4 inch specimen mix design

Table B-15 Compaction Data for 5% AC (G4 for Martin Marietta Gravel)
Densification Data @ 5.0 % Asphalt Content Gmm(meas) = 2.414 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 69.2 68.5 67.8 66.5 65.8 65.4 65.1 64.8 64.6 64.6 64.4 64.2 63.9 543.496 537.998 532.500 522.290 516.792 513.650 511.294 508.938 507.367 507.367 505.796 504.226 501.869 1129.6 2.301 1.022 2.078 2.100 2.121 2.163 2.186 2.199 2.209 2.220 2.226 2.226 2.233 2.240 2.251 2.125 2.146 2.169 2.211 2.235 2.248 2.259 2.269 2.276 2.276 2.283 2.290 2.301 88.0 88.9 89.8 91.6 92.6 93.1 93.6 94.0 94.3 94.3 94.6 94.9 95.3 68.9 68.3 67.7 66.5 65.8 65.3 65 64.7 64.6 64.5 64.3 64.1 63.9 541.139 536.427 531.715 522.290 516.792 512.865 510.509 508.153 507.367 506.582 505.011 503.440 501.869 1128.9 2.314 1.029 2.086 2.104 2.123 2.161 2.184 2.201 2.211 2.222 2.225 2.228 2.235 2.242 2.249 2.146 2.165 2.184 2.224 2.247 2.264 2.275 2.285 2.289 2.292 2.300 2.307 2.314 88.9 88.460 89.7 89.300 90.5 90.157 92.1 91.851 93.1 92.828 93.8 93.468 94.2 93.899 94.7 94.334 94.8 94.552 95.0 94.626 95.3 94.920 95.6 95.216 95.9 95.588 Specimen 2 Avg

204

Table B-16 Compaction Data for 5.5% AC (G4 for Martin Marietta Gravel)
Densification Data @ 5.5 % Asphalt Content Gmm(meas) = 2.398 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 69.9 69.5 69 67.6 66.9 66.4 66.1 65.8 65.7 65.5 65.4 65.2 65.1 548.993 545.852 541.925 530.929 525.431 521.504 519.148 516.792 516.007 514.436 513.650 512.080 511.294 1149.8 2.317 1.030 2.094 2.106 2.122 2.166 2.188 2.205 2.215 2.225 2.228 2.235 2.238 2.245 2.249 2.158 2.170 2.186 2.231 2.255 2.272 2.282 2.292 2.296 2.303 2.306 2.313 2.317 89.987 69.5 90.505 68.8 91.161 68.1 93.049 66.8 94.022 66 545.852 540.354 534.856 524.646 518.363 515.221 512.080 510.509 509.723 508.153 506.582 505.796 505.011 1144.95 2.323 1.025 2.098 2.119 2.141 2.182 2.209 2.222 2.236 2.243 2.246 2.253 2.260 2.264 2.267 2.149 2.171 2.193 2.236 2.263 2.277 2.291 2.298 2.302 2.309 2.316 2.319 2.323 89.624 89.806 90.536 90.521 91.467 91.314 93.247 93.148 94.377 94.200 94.953 94.842 95.535 95.348 95.829 95.712 95.977 95.858 96.273 96.153 96.572 96.375 96.722 96.598 96.872 96.747 Specimen 2 Avg

94.730 65.6 95.160 65.2 95.594 65

95.740 64.9 96.032 64.7 96.179 64.5 96.474 64.4 96.622 64.3

205

Table B-17 Compaction Data for 6% AC (G4 for Martin Marietta Gravel)
Densification Data @ 6.0 % Asphalt Content Gmm(meas) = 2.385 Specimen 1 Gyrations Ht, mm Volume, cm3 Gmb(est)Gmb(corr) %Gmm Ht, mmVolume, cm3 Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 69.5 68.8 68 66.7 66 65.5 65.2 64.9 64.8 64.7 64.5 64.4 64.3 545.852 540.354 534.071 523.861 518.363 514.436 512.080 509.723 508.938 508.153 506.582 505.796 505.011 1150.1 2.335 1.025 2.107 2.128 2.153 2.195 2.219 2.236 2.246 2.256 2.260 2.263 2.270 2.274 2.277 2.160 2.182 2.208 2.251 2.275 2.292 2.303 2.313 2.317 2.321 2.328 2.331 2.335 90.5784 69.1 91.5 68.4 542.710 537.212 531.715 522.290 516.792 513.650 511.294 508.938 508.153 507.367 505.796 505.011 504.226 1146 2.338 1.029 2.112 2.133 2.155 2.194 2.218 2.231 2.241 2.252 2.255 2.259 2.266 2.269 2.273 2.172 2.194 2.217 2.257 2.281 2.295 2.306 2.316 2.320 2.324 2.331 2.334 2.338 91.078 90.828 92.010 91.755 92.961 92.769 94.639 94.510 95.646 95.514 96.231 96.170 96.674 96.613 97.122 97.060 97.272 97.210 97.422 97.360 97.725 97.662 97.877 97.814 98.029 97.966 Specimen 2 Avg

92.5765 67.7 94.3808 66.5 95.3818 65.8 96.1099 65.4 96.5521 65.1 96.9984 64.8 97.1481 64.7 97.2983 64.6 97.6 64.4

97.7515 64.3 97.9036 64.2

206

Table B-18 Compaction Data for 6.5% AC (G4 for Martin Marietta Gravel)
Densification Data @ 6.5 % Asphalt Content Gmm(meas) = 2.369 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 69.5 68.8 68 66.7 66 65.5 65.2 64.9 64.8 64.7 64.5 64.3 64.2 545.852 540.354 534.071 523.861 518.363 514.436 512.080 509.723 508.938 508.153 506.582 505.011 504.226 1149.2 2.336 1.025 2.105 2.127 2.152 2.194 2.217 2.234 2.244 2.255 2.258 2.262 2.269 2.276 2.279 2.158 2.180 2.205 2.248 2.272 2.290 2.300 2.311 2.314 2.318 2.325 2.332 2.336 91.087 69.1 92.014 68.3 93.097 67.6 94.911 66.3 95.918 65.6 96.650 65.1 97.095 64.8 97.543 64.5 97.694 64.4 97.845 64.3 98.148 64.1 98.454 64 542.710 536.427 530.929 520.719 515.221 511.294 508.938 506.582 505.796 505.011 503.440 502.655 501.084 1151.65 2.344 1.020 2.122 2.147 2.169 2.212 2.235 2.252 2.263 2.273 2.277 2.280 2.288 2.291 2.298 2.164 2.190 2.212 2.256 2.280 2.297 2.308 2.319 2.322 2.326 2.333 2.337 2.344 91.356 91.221 92.426 92.220 93.383 93.240 95.214 95.062 96.230 96.074 96.969 96.809 97.418 97.256 97.871 97.707 98.023 97.858 98.175 98.010 98.482 98.315 98.636 98.545 98.945 98.776 Specimen 2 Avg

98.607 63.8

207

Table B-19 Summary of Compaction Data (G4 for Martin Marietta Gravel)
% Gmm, 6.0 Number of Gyration % Gmm, 5.0 AC % Gmm, 5.5 AC AC 5 7 10 20 30 40 50 60 65 70 80 90 99 88.460 89.300 90.157 91.851 92.828 93.468 93.899 94.334 94.552 94.626 94.920 95.216 95.588 89.806 90.521 91.314 93.148 94.200 94.842 95.348 95.712 95.858 96.153 96.375 96.598 96.747 90.828 91.755 92.769 94.510 95.514 96.170 96.613 97.060 97.210 97.360 97.662 97.814 97.966 91.221 92.220 93.240 95.062 96.074 96.809 97.256 97.707 97.858 98.010 98.315 98.545 98.776 % Gmm, 6.5 AC

Table B-20 Summary of Design Parameters (G4 for Martin Marietta Gravel)
% Gmm AC % 5 5.5 6 6.5 @N=7 89.300 90.521 91.755 92.220 % Gmm @N=65 94.552 95.858 97.210 97.858 % Gmm Air Void, %, @N=99 95.588 96.747 97.966 98.776 @ Ndes 5.448 4.142 2.790 2.142 Gsb 2.576 2.576 2.576 2.576 Ps 0.95 0.945 0.94 0.935 Gmm % VMA % VFA 2.426 2.413 2.391 2.383 15.41 15.15 15.19 15.36 64.64 72.65 81.63 86.06

208

Figure B-25 Gyratory 4 Mix Design (Gmm% Vs. Gyration No. for Martin Marietta Gravel)
Average Densification Curves
100 99 98 97 96 95 94

% Gmm

5.0 % AC 5.5 % AC 6.0 % AC 6.5 % AC

93 92 91 90 89 88 87 86 85
1 10 100 1000

Number of Gyrations

Figure B-26 Gyratory 4 Mix Design (Gmm% at Initial Vs. AC for Martin Marietta Gravel)
93

Percent of Gmm at Nini

92 92 91 91 90 90 89 4.5 5 5.5 6 6.5 7

Asphalt Content, %

209

Figure B-27 Gyratory 4 Mix Design (Gmm% at Max Vs. AC for Martin Marietta Gravel)
99

Percent of Gmm at N max

99 98 98 97 97 96 96 95 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-28 Gyratory 4 Mix Design (Air Void Vs. AC for Martin Marietta Gravel)
6

Air Voids, %

0 4.5 5 5.5 6 6.5 7

Asphalt Content, %

210

Figure B-29 Gyratory 4 Mix Design (VMA Vs. AC for Martin Marietta Gravel)
15.5 15.4 15.3

VMA, %

15.2 15.1 15.0 14.9 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-30 Gyratory 4 Mix Design (VFA Vs. AC for Martin Marietta Gravel)
90 85 80

VFA, %

75 70 65 60 55 50 4.5 5 5.5 6 6.5 7

Asphalt Content, %

211

 Gyratory 6 inch specimen mix design

Table B-21 Compaction Data for 4% AC (G6 for Martin Marietta Gravel)
Densification Data @ 4.0 % Asphalt Content Gmm(meas) = 2.452 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 125.2 123.9 122.5 120.4 119.2 118.4 117.8 117.4 117.2 117.1 116.8 116.5 116.3 2212.467 2189.494 2164.754 2127.644 2106.438 2092.301 2081.698 2074.629 2071.095 2069.328 2064.026 2058.725 2055.191 4627.7 2.298 1.021 2.092 2.114 2.138 2.175 2.197 2.212 2.223 2.231 2.234 2.236 2.242 2.248 2.252 2.135 2.157 2.182 2.220 2.242 2.257 2.269 2.276 2.280 2.282 2.288 2.294 2.298 87.1 88.0 89.0 90.5 91.4 92.1 92.5 92.8 93.0 93.1 93.3 93.6 93.7 125.9 124.7 123.5 121.2 120.1 119.3 118.7 118.3 118.1 118 117.6 117.4 117.2 2224.837 2203.631 2182.425 2141.781 2122.342 2108.205 2097.602 2090.534 2086.999 2085.232 2078.164 2074.629 2071.095 4645.5 2.298 1.025 2.088 2.108 2.129 2.169 2.189 2.204 2.215 2.222 2.226 2.228 2.235 2.239 2.243 2.139 2.160 2.181 2.222 2.243 2.258 2.269 2.277 2.280 2.282 2.290 2.294 2.298 87.2 87.150 88.1 88.027 88.9 88.957 90.6 90.577 91.5 91.448 92.1 92.063 92.5 92.531 92.8 92.845 93.0 93.002 93.1 93.082 93.4 93.359 93.6 93.559 93.7 93.719 Specimen 2 Avg

212

Table B-22 Compaction Data for 4.5% AC (G6 for Martin Marietta Gravel)
Densification Data @ 4.5 % Asphalt Content Gmm(meas) = 2.426 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 125.6 124.3 123.1 120.8 119.5 118.7 118.1 117.7 117.5 117.3 117 116.7 116.5 2219.535 2196.562 2175.357 2134.712 2111.739 2097.602 2086.999 2079.931 2076.396 2072.862 2067.561 2062.259 2058.725 4673.8 2.314 1.019 2.106 2.128 2.149 2.189 2.213 2.228 2.239 2.247 2.251 2.255 2.261 2.266 2.270 2.146 2.169 2.190 2.232 2.256 2.271 2.283 2.290 2.294 2.298 2.304 2.310 2.314 88.473 125.6 89.398 124.4 90.269 123.1 91.988 120.8 92.989 119.6 93.616 118.8 94.091 118.2 94.411 117.7 94.572 117.5 94.733 117.3 94.976 117 2219.535 2198.329 2175.357 2134.712 2113.506 2099.369 2088.766 2079.931 2076.396 2072.862 2067.561 2062.259 2058.725 4662 2.311 1.021 2.100 2.121 2.143 2.184 2.206 2.221 2.232 2.241 2.245 2.249 2.255 2.261 2.265 2.144 2.164 2.187 2.229 2.251 2.266 2.278 2.287 2.291 2.295 2.301 2.307 2.311 88.358 88.415 89.210 89.304 90.152 90.211 91.869 91.928 92.791 92.890 93.415 93.515 93.890 93.990 94.288 94.350 94.449 94.510 94.610 94.671 94.853 94.914 95.096 95.158 95.260 95.322 Specimen 2 Avg

95.220 116.7 95.383 116.5

213

Table B-23 Compaction Data for 5% AC (G6 for Martin Marietta Gravel)
Densification Data @ 5.0 % Asphalt Content Gmm(meas) = 2.414 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 125.5 124.2 122.8 120.4 119 118.2 117.5 117 116.8 116.6 116.3 116 115.7 2217.768 2194.795 2170.055 2127.644 2102.904 2088.766 2076.396 2067.561 2064.026 2060.492 2055.191 2049.889 2044.588 4667.8 2.328 1.020 2.105 2.127 2.151 2.194 2.220 2.235 2.248 2.258 2.262 2.265 2.271 2.277 2.283 2.146 2.169 2.193 2.237 2.263 2.279 2.292 2.302 2.306 2.310 2.316 2.322 2.328 88.907 125 2208.932 2187.727 2164.754 2122.342 2101.136 2086.999 2074.629 2067.561 2064.026 2060.492 2055.191 2049.889 2046.355 4680.2 2.332 1.020 2.119 2.139 2.162 2.205 2.227 2.243 2.256 2.264 2.268 2.271 2.277 2.283 2.287 2.160 2.181 2.204 2.249 2.271 2.287 2.300 2.308 2.312 2.316 2.322 2.328 2.332 89.493 89.200 90.361 90.099 91.320 91.091 93.144 92.909 94.084 93.924 94.722 94.560 95.287 95.123 95.612 95.489 95.776 95.653 95.940 95.817 96.188 96.064 96.437 96.312 96.603 96.520 Specimen 2 Avg

89.837 123.8 90.862 122.5 92.673 120.1 93.763 118.9 94.398 118.1 94.960 117.4 95.366 117

95.529 116.8 95.693 116.6 95.940 116.3 96.188 116

96.437 115.8

214

Table B-24 Compaction Data for 5.5% AC (G6 for Martin Marietta Gravel)
Densification Data @ 5.5 % Asphalt Content Gmm(meas) = 2.398 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 125.5 124.3 122.9 120.5 119.1 118.3 117.6 117.1 116.9 116.7 116.4 116.1 115.9 2217.768 2196.562 2171.822 2129.411 2104.671 2090.534 2078.164 2069.328 2065.794 2062.259 2056.958 2051.656 2048.122 4701.1 2.34 1.019 2.121 2.141 2.166 2.209 2.235 2.250 2.263 2.273 2.277 2.281 2.287 2.293 2.297 2.153 2.174 2.198 2.242 2.269 2.284 2.298 2.307 2.311 2.315 2.321 2.327 2.331 90.268 125.2 91.140 123.9 92.178 122.5 94.014 120 2212.467 2189.494 2164.754 2120.575 2095.835 2079.931 2069.328 2062.259 2056.958 2055.191 2048.122 2044.588 2039.286 4705.8 2.35 1.016 2.132 2.154 2.179 2.224 2.250 2.268 2.279 2.287 2.293 2.295 2.303 2.307 2.313 2.165 2.188 2.213 2.259 2.285 2.303 2.315 2.323 2.329 2.331 2.339 2.343 2.349 90.775 90.521 91.727 91.433 92.775 92.477 94.708 94.361 95.826 95.473 96.559 96.161 97.054 96.693 97.386 97.065 97.637 97.273 97.721 97.398 98.059 97.692 98.228 97.902 98.483 98.114 Specimen 2 Avg

95.119 118.6 95.762 117.7 96.332 117.1 96.744 116.7 96.909 116.4 97.075 116.3 97.325 115.9 97.577 115.7 97.745 115.4

215

Table B-25 Compaction Data for 6% AC (G6 for Martin Marietta Gravel)
Densification Data @ 6.0 % Asphalt Content Gmm(meas) = 2.385 Specimen 1 Gyrations Ht, mm Volume, cm3Gmb(est)Gmb(corr)%GmmHt, mmVolume, cm3Gmb(est) Gmb(corr)%Gmm%Gmm 5 7 10 20 30 40 50 60 65 70 80 90 99 Weight (gm) = Gmb(meas)= CF= 123.2 122.2 120.9 118.4 117.1 116.3 115.7 115.2 115 114.8 114.5 114.2 114.1 2177.124 2159.452 2136.479 2092.301 2069.328 2055.191 2044.588 2035.752 2032.218 2028.683 2023.382 2018.081 2016.313 4703.9 2.368 1.015 2.161 2.178 2.202 2.248 2.273 2.289 2.301 2.311 2.315 2.319 2.325 2.331 2.333 2.193 2.211 2.235 2.282 2.307 2.323 2.335 2.345 2.349 2.354 2.360 2.366 2.368 91.953 124.7 92.706 123.4 93.703 122.1 95.681 119.6 96.744 118.2 97.409 117.3 97.914 116.7 98.339 116.2 98.510 116 2203.631 2180.658 2157.685 2113.506 2088.766 2072.862 2062.259 2053.423 2049.889 2046.355 2041.053 2035.752 2032.218 4716.4 2.357 1.016 2.140 2.163 2.186 2.232 2.258 2.275 2.287 2.297 2.301 2.305 2.311 2.317 2.321 2.174 2.197 2.220 2.266 2.293 2.311 2.323 2.333 2.337 2.341 2.347 2.353 2.357 91.139 91.546 92.099 92.402 93.079 93.391 95.025 95.353 96.151 96.447 96.888 97.149 97.386 97.650 97.805 98.072 97.974 98.242 98.143 98.413 98.398 98.669 98.654 98.927 98.826 99.057 Specimen 2 Avg

98.682 115.8 98.940 115.5 99.200 115.2 99.287 115

216

Table B-26 Summary of Compaction Data (G6 for Martin Marietta Gravel)
Number of % Gmm, 4.0 AC % Gmm, 4.5 AC % Gmm, 5.0 AC % Gmm, 5.5 AC % Gmm, 6.0 AC Gyration 5 7 10 20 30 40 50 60 65 70 80 90 99 87.150 88.027 88.957 90.577 91.448 92.063 92.531 92.845 93.002 93.082 93.359 93.559 93.719 88.415 89.304 90.211 91.928 92.890 93.515 93.990 94.350 94.510 94.671 94.914 95.158 95.322 89.200 90.099 91.091 92.909 93.924 94.560 95.123 95.489 95.653 95.817 96.064 96.312 96.520 90.222 91.131 92.171 94.049 95.157 95.842 96.373 96.744 96.951 97.076 97.369 97.579 97.790 91.546 92.402 93.391 95.353 96.447 97.149 97.650 98.072 98.242 98.413 98.669 98.927 99.057

Table B-27 Summary of Design Parameters (G6 for Martin Marietta Gravel)
% Gmm AC % @N=7 4 4.5 5 5.5 6 88.027 89.304 90.099 91.433 92.402 @N=65 93.002 94.510 95.653 97.273 98.242 @N=99 93.719 95.322 96.520 98.114 99.057 @ Ndes 6.998 5.490 4.347 2.727 1.758 2.576 2.576 2.576 2.576 2.576 0.96 0.955 0.95 0.945 0.94 2.452 2.426 2.414 2.398 2.379 15.015 53.398 14.998 63.398 14.845 70.714 14.429 81.102 14.715 88.054 % Gmm % Gmm Air Void, %, Gsb Ps Gmm % VMA % VFA

217

Figure B-31 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Martin Marietta Gravel)
Average Densification Curves
100 99 98 97 96 95 94 4.0 % AC 4.5% AC 5.0% AC 5.5% AC 6.0% AC

% Gmm

93 92 91 90 89 88 87 86 85 1 10 100 1000

Number of Gyrations

Figure B-32 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Martin Marietta Gravel)
93.0 92.5

Percent of Gmm at Nini

92.0 91.5 91.0 90.5 90.0 89.5 89.0 88.5 88.0 87.5 3.4 3.9 4.4 4.9 5.4 5.9 6.4

Asphalt Content, %

218

Figure B-33 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Martin Marietta Gravel)
99.5 99.0 98.5 98.0 97.5 97.0 96.5 96.0 95.5 95.0 94.5 94.0 93.5 93.0 3.5 4 4.5 5 5.5 6 6.5

P ercent of Gmm at N max

Asphalt Content, %

Figure B-34 Gyratory 6 Mix Design (Air Void Vs. AC for Martin Marietta Gravel)
8 7 6

A ir V o id s , %

5 4 3 2 1 0 3.5 4 4.5 5 5.5 6 6.5

Asphalt Content, %

219

Figure B-35 Gyratory 6 Mix Design (VMA Vs. AC for Martin Marietta Gravel)
15.5 15.4 15.3 15.2 15.1

VMA, %

15.0 14.9 14.8 14.7 14.6 14.5 14.4 14.3 3.5 4 4.5 5 5.5 6 6.5

Asphalt Content, %

Figure B-36 Gyratory 6 Mix Design (VFA Vs. AC for Martin Marietta Gravel)
95 90 85 80

VFA, %

75 70 65 60 55 50 3.6 4.1 4.6 5.1 5.6 6.1 6.6

Asphalt Content, %

220

B3. Traprock Source

Marshall mix design

Table B-28 Marshall mix design for Ontario Traprock

Traprock % AC by wt. Unit Weight, % Air Voids of mix. 4.5 5 5.5 6 6.5 pcf (Mg/m3) 153.75 155.94 157.50 158.37 159.18 9.38 6.58 4.54 3.86 2.41 14.20 14.25 14.51 14.49 14.51 40.11 55.87 68.72 73.34 83.39 % VMA % VFA (N) 2291 2087 2109 1948 1987 (0.25 mm) 10.3 11 11.4 11.8 12.1 Stability, lbs Flow 0.01 in.

221

Figure B-37 Marshall Mix Design (AV Vs. AC for Ontario Traprock)
Air Voids, %
10.0 9.0 8.0

Optimum AC=5.8% @ 4.0 % Air voids

Air Voids, %

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-38 Marshall Mix Design (VMA Vs. AC for Ontario Traprock)
VM A, %
15 14.8 14.6

VMA, %

14.4 14.2 14 13.8 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

222

Figure B-39 Marshall Mix Design (Stability Vs. AC for Ontario Traprock)
Stability, pounds 2400 2350 2300 Stability, Pounds 2250 2200 2150 2100 2050 2000 1950 1900 3.5 4 4.5 5 5.5 6 6.5 7 Asphalt Content, %

Figure B-40 Marshall Mix Design (Flow Vs. AC for Ontario Traprock)
Flow, 0.01 in
16

14

Flow, 0.01 in

12

10

8 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

223

Figure B-41 Marshall Mix Design (VFA Vs. AC for Ontario Traprock)
VFA, %
90.0 85.0 80.0 75.0 70.0 65.0 60.0 55.0 50.0 45.0 40.0 35.0 30.0 3.5 4 4.5 5 5.5 6 6.5 7

VFA, %

Asphalt Content, %

Figure B-42 Marshall Mix Design (Unit WGT Vs. AC for Ontario Traprock)
Unit Weight, pcf
160.00

Unit weight, pcf

159.00 158.00 157.00 156.00 155.00 154.00 153.00 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

224

 Gyratory 6 inch mix design

Table B-29 Summary of Design Parameters (G6 for Ontario Traprock)

%Gmm AC % @N=7 4.5 5 5.5 6 87.163 87.815 88.779 89.506 @N=65 93.990 94.843 96.130 97.278 @N=99 95.322 96.361 97.546 98.808 %,@Ndes 6.010 5.157 3.870 2.722 2.576 2.576 2.576 2.576 0.955 0.95 0.945 0.94 2.426 2.418 2.404 2.391 15.466 15.426 15.223 15.126 61.140 66.567 74.577 82.004 %Gmm %Gmm Air Void, Gsb Ps Gmm %VMA %VFA

Figure B-43 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Ontario Traprock)
Average Densification Curves
100.000 99.000 98.000 97.000 96.000 95.000 94.000 93.000 92.000 91.000 90.000 89.000 88.000 87.000 86.000 85.000 84.000 83.000 1 10 100 1000
4.0 % AC 4.5% AC 5.0% AC 5.5% AC 6.0% AC

% Gmm

Number of Gyrations

225

Figure B-44 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Ontario Traprock)
90.000 89.500

Percent of Gmm at Nini

89.000 88.500 88.000 87.500 87.000 86.500 3.4 3.9 4.4 4.9 5.4 5.9 6.4

Asphalt Content, %

Figure B-45 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Ontario Traprock)
99.000 98.500 Percent of Gmm at Nmax 98.000 97.500 97.000 96.500 96.000 95.500 95.000 3.5 4 4.5 5 5.5 6 6.5 Asphalt Content, %

226

Figure B-46 Gyratory 6 Mix Design (Air Void Vs. AC for Ontario Traprock)
7.000

6.000

5.000

Air Voids, %

4.000

3.000

2.000

1.000

0.000 3.5 4 4.5 5 5.5 6 6.5

Asphalt Content, %

Figure B-47 Gyratory 6 Mix Design (VMA Vs. AC for Ontario Traprock)
15.500

15.400

VM A, %

15.300

15.200

15.100 3.5 4 4.5 5 5.5 6 6.5

Asphalt Content, %

227

Figure B-48 Gyratory 6 Mix Design (Air Void Vs. AC for Ontario Traprock)
85 80 75

VFA, %

70 65 60 55 50 3.6 4.1 4.6 5.1 5.6 6.1 6.6

Asphalt Content, %

228

B4. Stocker Sand Gravel

Marshall mix design

Table B-30 Marshall mix design for Stocker Sand Gravel

Stocker Gravel % AC by wt. of mix. 4 4.5 5 5.5 6 6.5 Unit Weight, pcf (Mg/m3) 140.75 140.37 142.38 145.53 145.55 145.70 % Air %VMA Voids 10.70 10.16 7.72 5.08 4.05 3.51 17.98 18.63 17.89 16.52 16.95 17.30 40.46 45.43 56.86 69.26 76.10 79.70 %VFA lbs(N) 1786 1797 1665 1567 1607 1534 7.4 7.8 8.1 8.7 9.4 10.7 Stability, Flow 0.01 in.

229

Figure B-49 Marshall Mix Design (AV Vs. AC for Stocker Sand Gravel)
Air Voids, %
12.0 11.0 10.0 9.0 Optim um AC=6% @ 4.0 % Air voids

Air Voids, %

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-50 Marshall Mix Design (VMA Vs. AC for Stocker Sand Gravel)
VMA, %
19.00 18.50 18.00

VMA, %

17.50 17.00 16.50 16.00 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

230

Figure B-51 Marshall Mix Design (Stability Vs. AC for Stocker Sand Gravel)
Stability, pounds
1800 1750

Stability, Pounds

1700 1650 1600 1550 1500 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

Figure B-52 Marshall Mix Design (Flow Vs. AC for Stocker Sand Gravel)
Flow, 0.01 in
16 14

Flow, 0.01 in

12 10 8 6 3.5 4 4.5 5 5.5 6 6.5 7

Asphalt Content, %

231

Figure B-53 Marshall Mix Design (VFA Vs. AC for Stocker Sand Gravel)
VFA, %
85.0 80.0 75.0 70.0 65.0 60.0 55.0 50.0 45.0 40.0 35.0 30.0 3.5 4 4.5 5 5.5 6 6.5 7

VFA, %

Asphalt Content, %

Figure B-54 Marshall Mix Design (Unit WGT Vs. AC for Stocker Sand Gravel)
Unit Weight, pcf

148.00

Unit weight, pcf

146.00 144.00 142.00 140.00 138.00 4 4.5 5 5.5 6 6.5 7 7.5

Asphalt Content, %

232

 Gyratory 6 inch specimen mix design

Table B-31 Summary of Design Parameters (G6 for Stocker Sand Gravel)
Air Void, % Gmm AC % @N=7 @N=65 @N=99 Ndes 4 4.5 5 5.5 6 86.085 87.163 87.815 88.779 89.506 92.413 93.990 94.843 96.130 97.278 93.719 95.322 96.361 97.546 98.808 7.587 6.010 5.157 3.870 2.722 2.576 2.576 2.576 2.576 2.576 0.96 0.955 0.95 0.945 0.94 2.452 2.426 2.418 2.404 2.391 15.554 15.466 15.426 15.223 15.126 51.220 61.140 66.567 74.577 82.004 % Gmm % Gmm %, @ Gsb Ps Gmm %VMA %VFA

233

Figure B-55 Gyratory 6 Mix Design (Gmm% Vs. Gyration No. for Stocker Sand Gravel)

Average Densification Curves

100.000 99.000 98.000 97.000 96.000 95.000 94.000
4.0 % AC 4.5% AC 5.0% AC 5.5% AC 6.0% AC

% Gmm

93.000 92.000 91.000 90.000 89.000 88.000 87.000 86.000 85.000 84.000 83.000 1 10 100 1000

Number of Gyrations

234

Figure B-56 Gyratory 6 Mix Design (Gmm% at Initial Vs. AC for Stocker Sand Gravel)

90.000 89.500

Percent of Gmm at Nini

89.000 88.500 88.000 87.500 87.000 86.500 86.000 85.500 3.4 3.9 4.4 4.9 5.4 5.9 6.4

Asphalt Content, %

Figure B-57 Gyratory 6 Mix Design (Gmm% at Max Vs. AC for Stocker Sand Gravel)
99.500 99.000 98.500

Percent of Gmm at Nmax

98.000 97.500 97.000 96.500 96.000 95.500 95.000 94.500 94.000 93.500 93.000 3.5 4 4.5 5 5.5 6 6.5

Asphalt Content, %

235

Figure B-58 Gyratory 6 Mix Design (Air Void Vs. AC for Stocker Sand Gravel)
8.000

7.000

6.000

A ir V o ids, %

5.000

4.000

3.000

2.000

1.000

0.000 3.5 4 4.5 5 5.5 6 6.5

Asphalt Conte nt, %

Figure B-59 Gyratory 6 Mix Design (VMA Vs. AC for Stocker Sand Gravel)
15.500

15.400

V MA , %

15.300

15.200

15.100 3.5 4 4.5 5 5.5 6 6.5

Asphalt Content, %

236

Figure B-60 Gyratory 6 Mix Design (VFA Vs. AC for Stocker Sand Gravel)
85 80 75

VFA, %

70 65 60 55 50 3.6 4.1 4.6 5.1 5.6 6.1 6.6

Asphalt Content, %

237