VERIFICATION OF THE SUPERPAVE GYRATORY Ndesign COMPACTION LEVELS Except where reference is made to the work of others, the work described in this dissertation is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. ____________________________________ Brian Douglas Prowell Certificate of Approval: _____________________________ __________________________ Frazier Parker, Jr. E. Ray Brown, Chair Professor Professor Civil Engineering Civil Engineering _____________________________ __________________________ David H. Timm Stephen L. McFarland Assistant Professor Acting Dean Civil Engineering Graduate School VERIFICATION OF THE SUPERPAVE GYRATORY Ndesign COMPACTION LEVELS Brian Douglas Prowell A Dissertation Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirement for the Degree of Doctorate of Philosophy Auburn, Alabama May 11, 2006 iii VERIFICATION OF THE SUPERPAVE GYRATORY Ndesign COMPACTION LEVELS Brian Douglas Prowell Permission is granted to Auburn University to make copies of this dissertation at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. ______________________________ Signature of Author ______________________________ Date of Graduation iv VITA Brian Douglas Prowell, son of Harry Douglas and Judy Prowell was born February 12, 1967 in Harrisburg, Pennsylvania. He graduated from Mechanicsburg Area Senior High School in May 1985. He entered George Washington University in August 1985. After a rocky start where he found a greater passion for rowing than for engineering, he entered the Pennsylvania State University in 1987, and graduated with a Bachelor of Science in Civil Engineering in August 1990. He then worked for Kidde Consultants as an engineering intern from August 1990 until December 1990. In January 1991, he entered Graduate School at Virginia Polytechnic Institute and State University, and graduated with a Master of Science degree in Civil Engineering in August 1992. In August 1992, he married Marcia Ann Votour, the daughter of Paul and Carol Votour. From August 1992 until May 1993 he was employed as an instructor in the Civil Engineering Department at Virginia Polytechnic Institute and State University. He was then employed as a research scientist and later senior research scientist by the Virginia Transportation Research Council, the research arm of the Virginia Department of Transportation from May 1993 until July 2001. He has been employed with the National Center for Asphalt Technology at Auburn University since August 2001 as a research engineer and later as an assistant director. Marcia and Brian have been blessed with two children since moving to Auburn, Alabama, Benjamin Thomas Prowell, born September 25, 2001, and Katherine Elizabeth Prowell, born May 13, 2004. v DISSERTATION ABSTRACT VERIFICATION OF THE SUPERPAVE GYRATORY Ndesign COMPACTION LEVELS Brian Douglas Prowell Doctor of Philosophy, May 11, 2006 (M. S., Virginia Polytechnic Institute and State University, 1992) (B. S., Pennsylvania State University, 1990) 293 Typed Pages Directed by E. Ray Brown The original Superpave Ndesign table contained 28 levels based on a limited laboratory experiment. This was later consolidated to 4 levels based on the sensitivity of mixture volumetric properties to Ndesign; however, these data were not verified as being correct for field conditions. An experiment was conducted to verify the Ndesign levels in the field. Samples were collected, tested and analyzed from 40 field projects. The projects were selected in a total of 16 states. The projects represent a wide range of traffic levels, binder grades, aggregate types, and gradations. Each project was visited at the time of construction and at 5 additional times after construction. The 40 pavements vi studied in this project appeared to reach their ultimate density after two years of traffic. A fair relationship was determined between the as-constructed and the density after two years of traffic. The high temperature PG binder grade was found to significantly affect pavement densification, with stiffer binders resulting in less densification. The ultimate in-place densities of the pavements evaluated in this study were approximately 1.5 percent less than the densities of the laboratory compacted samples at the agency specified Ndesign. The number of gyrations to match the ultimate in-place density was calculated for each project in this study. The calculated values for the two compactors used in this study differed by approximately 20 gyrations. This was attributed to differences in their dynamic internal angle. The predicted gyrations, adjusted to a dynamic internal angle of 1.16 degrees showed good agreement between the two machines. A relationship was developed between predicted Ndesign and design traffic for the projects which were not constructed using PG 76-22. Although there was a great deal of scatter in the data, the scatter was expected. The predicted gyration levels were generally less than those currently specified. All of the projects in this study were very rut resistant. The maximum observed rutting for the field projects was 7.4 mm with an average rut depth for all of the projects of 2.7 mm after 4 years of traffic. Based on the densification and performance data the Ndesign levels can be reduced for higher traffic levels and the Ninitial and Nmaximum criteria can be eliminated. vii AKNOWLEDGEMENTS The author thanks all of the state agencies and contractors who assisted with this project. The author thanks Shane Buchanan, Mike Huner, Graham Hurley, Robert James, Jason Moore, and all of the staff of the National Center for Asphalt Technology who began this project and assisted in the early field data collection. The author thanks Dr. Ray Brown for his guidance with this and many other projects. Dr. Brown?s knowledge and experience in hot mix asphalt pavements have greatly benefited the author. The author thanks Dr. David Timm and Dr. Frazier Parker for their thoughtful review of this work. The author thanks Dr. Saeed Maghsoodloo for his assistance with statistical analyses. The author thanks his parents for their encouragement throughout his academic career. The author especially thanks his wife, Marcia for her love, support, and patience and his children Ben and Kate for smiles and laughter. viii Style manual or journal used: National Cooperative Highway Research Council Computer software used: Microsoft Word 2002, Microsoft Excel 2002, and MINITAB? Release 14.13 ix TABLE OF CONTENTS LIST OF TABLES............................................................................................................ xii LIST OF FIGURES ....................................................................................................... xviii CHAPTER 1 INTRODUCTION AND RESEARCH APPROACH ...................................1 1.1 BACKGROUND...................................................................................................1 1.2 RESEARCH PROBLEM STATEMENT..............................................................2 1.3 OBJECTIVE..........................................................................................................3 1.4 SCOPE...................................................................................................................4 CHAPTER 2 LITERATURE REVIEW ..............................................................................5 2.1 A BRIEF HISTORY OF HMA MIX DESIGN PRIOR TO SUPERPAVE .......................................................................................................5 2.1.1 Proprietary Mixes.........................................................................................5 2.1.2 Hubbard-Field..............................................................................................9 2.1.3 Hveem Method.............................................................................................9 2.1.4 Marshall Mix Design .................................................................................13 2.1.5 Texas Gyratory Method .............................................................................23 2.1.6 Corps of Engineers Gyratory Compactor ..................................................25 2.1.7 French Design Procedure...........................................................................29 2.2 SUPERPAVE GYRATORY COMPACTOR .....................................................32 x 2.2.1 Selection of the SGC for the Superpave Mix Design System ...................32 2.2.2 Studies to Evaluate Factors Affecting Gyratory Compaction....................40 2.2.3 Internal Angle of Gyration.........................................................................46 2.3 DENSIFICATION OF PAVEMENTS UNDER TRAFFIC ...............................53 2.4 STUDIES RELATED TO Ndesign.....................................................................63 2.4.1 Development of the Original Ndesign Table..................................................63 2.4.2 Research Related to Ndesign Conducted after SHRP ....................................70 2.5 SUMMARY OF LITERATURE REVIEW ........................................................88 CHAPTER 3 RESEARCH TEST PLAN ..........................................................................93 3.1 RESEARCH TEST PLAN ..................................................................................93 CHAPTER 4 TEST RESULTS AND ANALYSES..........................................................98 4.1 PROJECTS SELECTED.....................................................................................98 4.2 TEST RESULTS ...............................................................................................101 4.2.1 Comparison of Mixture Data to Design Job Mix Formula...........................102 4.2.2 Estimation of Traffic.....................................................................................106 4.2.3 Pavement Densification................................................................................113 4.2.4 Determination of Ndesign to Match Ultimate In-Place Density ..................130 4.2.5 Evaluation of Locking Point.........................................................................161 4.2.6 Pavement Condition after Four Years ..........................................................165 4.2.7 Evaluation of Ninitial ...................................................................................167 4.2.8 Evaluation of Nmaximum.............................................................................169 4.2.9 Summary and Discussion of Test Results ....................................................171 CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS....................................179 xi 5.1 CONCLUSIONS ...............................................................................................180 5.2 RECOMMENDATIONS ..................................................................................182 CHAPTER 6 REFERENCES ..........................................................................................183 APPENDIX ..................................................................................................................192 xii LIST OF TABLES TABLE 2.1 Comparison of Densification Parameters from Gyratory Compactors.........................................................................................................42 TABLE 2.2 Experimental Matrix for Original Ndesign Experiment ................................64 TABLE 2.3 Ndesign Models .............................................................................................67 TABLE 2.4 Original Ndesign Table..................................................................................69 TABLE 2.5 Comparison of Ndesign Levels for Hot Climate for 1 and 1.3 Degrees.........................................................................................................76 TABLE 2.6 Revised Ndesign Table Proposed by Buchanan ............................................80 TABLE 2.7 Superpave Gyratory Compaction Effort ........................................................82 TABLE 2.8 Sample Gyratory Height Data Illustrating Locking Point Determination...........................................................................................................87 TABLE 3.1 Test Plan for Field Densification Study.........................................................94 TABLE 4.1 Summary of Project Information ...................................................................99 TABLE 4.2 Design Gradation and Optimum Asphalt Content (JMF) ............................103 TABLE 4.3 Lane Distribution Factors.............................................................................107 TABLE 4.4 Factors used to Calculate Accumulated ESALs at Various Intervals..............................................................................................................109 TABLE 4.5 Accumulated ESALs at Sampling Intervals.................................................111 TABLE 4.6 Average In-Place Densities for Field Projects .............................................114 xiii TABLE 4.7 Comparison of 2-Year and 4-Year Densities...............................................120 TABLE 4.8 ANOVA (GLM) Results for 3-Month Densification...................................125 TABLE 4.9 Original and Adjusted Gyrations to Match In-Place Density at 2 Years............................................................................................................135 TABLE 4.10 Predicted Gyrations to Match Ultimate Density........................................145 TABLE 4.11 Predicted Gyrations to Match In-Place Density.........................................147 TABLE 4.12 Matrix of Predicted Percentage of Laboratory Density .............................154 TABLE 4.13 Pine Predicted Gyration to Match Percentage of Lab Density ..................155 TABLE 4.14 Troxler Predicted Gyration to Match Percentage of Lab Density .............157 TABLE 4.15 Matrix of Gyrations....................................................................................160 TABLE 4.16 Four-Year Rut Depth Measurements .........................................................166 TABLE 4.17 Summary of Densities at Ninitial...............................................................168 TABLE 4.18 Summary of Densities at Nmaximum........................................................170 TABLE 4.19 Proposed Ndesign Levels for an SGC DIA of 1.16 ? 0.02 Degrees ............................................................................................................................178 TABLE 5.1 Recommended Ndesign Levels for an SGC DIA of 1.16 ? 0.02 Degrees ............................................................................................................................182 TABLE A.1 SGC Data for Project AL-1.........................................................................193 TABLE A.2 SGC Data for Project AL-2.........................................................................194 TABLE A.3 SGC Data for Project AL-3.........................................................................195 TABLE A.4 SGC Data for Project AL-4.........................................................................196 TABLE A.5 SGC Data for Project AL-5.........................................................................197 TABLE A.6 SGC Data for Project AL-6.........................................................................198 xiv TABLE A.7 SGC Data for Project AR-1.........................................................................199 TABLE A.8 SGC Data for Project AR-2.........................................................................200 TABLE A.9 SGC Data for Project AR-3.........................................................................201 TABLE A.10 SGC Data for Project AR-4.......................................................................202 TABLE A.11 SGC Data for Project CO-1.......................................................................203 TABLE A.12 SGC Data for Project CO-2.......................................................................204 TABLE A.13 SGC Data for Project CO-3.......................................................................205 TABLE A.14 SGC Data for Project CO-4.......................................................................206 TABLE A.15 SGC Data for Project CO-5.......................................................................207 TABLE A.16 SGC Data for Project FL-1 .......................................................................208 TABLE A.17 SGC Data for Project GA-1 ......................................................................209 TABLE A.18 SGC Data for Project IL-1 ........................................................................210 TABLE A.19 SGC Data for Project IL-2 ........................................................................211 TABLE A.20 SGC Data for Project IL-3 ........................................................................212 TABLE A.21 SGC Data for Project IN-1........................................................................213 TABLE A.22 SGC Data for Project IN-2........................................................................214 TABLE A.23 SGC Data for Project KS-1.......................................................................215 TABLE A.24 SGC Data for Project KY-1 ......................................................................216 TABLE A.25 SGC Data for Project KY-2 ......................................................................217 TABLE A.26 SGC Data for Project KY-3 ......................................................................218 TABLE A.27 SGC Data for Project MI-1 .......................................................................219 TABLE A.28 SGC Data for Project MI-2 .......................................................................220 TABLE A.29 SGC Data for Project MI-3 .......................................................................221 xv TABLE A.30 SGC Data for Project MO-1......................................................................222 TABLE A.31 SGC Data for Project MO-2......................................................................223 TABLE A.32 SGC Data for Project MO-3......................................................................224 TABLE A.33 SGC Data for Project NC-1.......................................................................225 TABLE A.34 SGC Data for Project NE-1.......................................................................226 TABLE A.35 SGC Data for Project NE-2.......................................................................227 TABLE A.36 SGC Data for Project NE-3.......................................................................228 TABLE A.37 SGC Data for Project NE-4.......................................................................229 TABLE A.38 SGC Data for Project TN-1.......................................................................230 TABLE A.39 SGC Data for Project UT-1.......................................................................231 TABLE A.40 SGC Data for Project WI-1.......................................................................232 TABLE A.41 Core Data for Project AL-1.......................................................................233 TABLE A.42 Core Data for Project AL-2.......................................................................234 TABLE A.43 Core Data for Project AL-3.......................................................................235 TABLE A.44 Core Data for Project AL-4.......................................................................236 TABLE A.45 Core Data for Project AL-5.......................................................................237 TABLE A.46 Core Data for Project AL-6.......................................................................238 TABLE A.47 Core Data for Project AR-1.......................................................................239 TABLE A.48 Core Data for Project AR-2.......................................................................240 TABLE A.49 Core Data for Project AR-3.......................................................................241 TABLE A.50 Core Data for Project AR-4.......................................................................242 TABLE A.51 Core Data for Project CO-1.......................................................................243 TABLE A.52 Core Data for Project CO-2.......................................................................244 xvi TABLE A.53 Core Data for Project CO-3.......................................................................245 TABLE A.54 Core Data for Project CO-4.......................................................................246 TABLE A.55 Core Data for Project CO-5.......................................................................247 TABLE A.56 Core Data for Project FL-1 .......................................................................248 TABLE A.57 Core Data for Project GA-1 ......................................................................249 TABLE A.58 Core Data for Project IL-1 ........................................................................250 TABLE A.59 Core Data for Project IL-2 ........................................................................251 TABLE A.60 Core Data for Project IL-3 ........................................................................252 TABLE A.61 Core Data for Project IN-1........................................................................253 TABLE A.62 Core Data for Project IN-2........................................................................254 TABLE A.63 Core Data for Project KS-1.......................................................................255 TABLE A.64 Core Data for Project KY-1 ......................................................................256 TABLE A.65 Core Data for Project KY-2 ......................................................................257 TABLE A.66 Core Data for Project KY-3 ......................................................................258 TABLE A.67 Core Data for Project MI-1 .......................................................................259 TABLE A.68 Core Data for Project MI-2 .......................................................................260 TABLE A.69 Core Data for Project MI-3 .......................................................................261 TABLE A.70 Core Data for Project MO-1......................................................................262 TABLE A.71 Core Data for Project MO-2......................................................................263 TABLE A.72 Core Data for Project MO-3......................................................................264 TABLE A.73 Core Data for Project NC-1.......................................................................265 TABLE A.74 Core Data for Project NE-1.......................................................................266 TABLE A.75 Core Data for Project NE-2.......................................................................267 xvii TABLE A.76 Core Data for Project NE-3.......................................................................268 TABLE A.77 Core Data for Project NE-4.......................................................................269 TABLE A.78 Core Data for Project TN-1.......................................................................270 TABLE A.79 Core Data for Project UT-1.......................................................................271 TABLE A.80 Core Data for Project WI-1.......................................................................272 xviii LIST OF FIGURES Figure 2.1. Stability and Durability as a Function of Asphalt Content..............................13 Figure 2.2. Model C Tournpull with Specially Built Loading Cart...................................18 Figure 2.3. Traffic Compaction Data for Mix 11, Crushed Limestone with Medium Filler Content.......................................................................................................19 Figure 2.4. Manual Texas Gyratory Molding Machine .....................................................24 Figure 2.5. Schematic of Compaction Head for Corps of Engineers Gyratory Compactor ..........................................................................................................26 Figure 2.6. Comparison of Laboratory and Field Density and Stability Values................27 Figure 2.7. Aggregate Density as a Function of Asphalt Content and Compaction Level..............................................................................................................29 Figure 2.8. Compaction Principle of the PCG ...................................................................30 Figure 2.9. Typical Gyratory Compaction Curve ..............................................................41 Figure 2.10. Effect of Asphalt Content on Compaction ....................................................41 Figure 2.11. DAVK and Calibration Block .......................................................................49 Figure 2.12. Definition of Internal and External Angle of Gyration .................................49 Figure 2.13. Gmb versus Average Internal Angle of Gyration..........................................52 Figure 2.14. 50-Blow Marshall versus In-Place ................................................................55 Figure 2.15. Densification as a Function of Initial Density...............................................60 xix Figure 2.16. Densification versus ESALs..........................................................................61 Figure 2.17. Comparison of Ndesign from Angles of 1 and 1.3 Degrees..........................66 Figure 2.18. Variation in VMA with Ndesign for PG 64-22.............................................72 Figure 2.19. Error in Back Calculated Air Voids Versus Gyration Level.........................74 Figure 2.20. Comparison of Ndesign and In-Place Air Voids after 3 Years .....................84 Figure 2.21. Comparison of Superpave and Marshall in-place Air Voids.........................85 Figure 2.22. Comparison of Superpave and Marshall Design VMA.................................86 Figure 3.1. Location of Field Projects ...............................................................................94 Figure 4.1. Frequency Distribution of Lift Thickness to NMAS by Gradation...............100 Figure 4.2. Design versus Average Field Percent Passing the 2.36 mm Sieve................104 Figure 4.3. Design versus Average Field Percent Passing the 0.075 mm Sieve..............104 Figure 4.4. Design versus Average Field Asphalt Content..............................................105 Figure 4.5. Distribution of 20-Year Design Traffic.........................................................112 Figure 4.6. Cumulative Frequency Distribution of As-Constructed, In-place Density...............................................................................................................115 Figure 4.7. Main Effect Plot for Factors Affecting As-Constructed Density ..................117 Figure 4.8. Cumulative Frequency Plot for In-Place Density by Sampling Period...............................................................................................................118 Figure 4.9. Densification of Project CO-4 with Time and Traffic...................................122 Figure 4.10. Densification of Project AL-1 with Time and Traffic.................................123 Figure 4.11. Densification of Project MI-1 with Time and Traffic .................................123 Figure 4.12. Main Effects Plot for Factors Effecting 3 Month Densification ................125 xx Figure 4.13. Main Effect Plot for Month of Construction on 2-Year Densification....................................................................................................................127 Figure 4.14. Average Test Track Pavement Densification..............................................129 Figure 4.15. Predicted Gyrations to Match Ultimate Density .........................................132 Figure 4.16. Comparison of Predicted Gyrations to match In-Place Density after Two-Years with and without Correction for DIA ...................................................134 Figure 4.17. Predicted Gyrations to Match Two-Year Density Corrected to a DIA of 1.16 Degrees ................................................................................................................136 Figure 4.18. Predicted Gyrations to Match In-Place Density Corrected to a DIA of 1.16 Degrees ...............................................................................................................137 Figure 4.19. In-Place 2 Year versus Agency Specified Ndesign Air Voids....................138 Figure 4.20. Average Gyrations to Match 2000 NCAT Test Track Density...................140 Figure 4.21. Predicted Gyrations to Match 2000 NCAT Test Track Density ................141 Figure 4.22. Predicted Gyrations for Pine SGC Excluding Projects Using PG 76-22................................................................................................................142 Figure 4.23. Standardized Residuals versus Fitted Mean for Log Predicted Gyrations versus Log 20-Year ESALs ............................................................................144 Figure 4.24. Predicted Gyrations versus 20 Year Design Traffic without PG 76-22 Data..................................................................................................................144 Figure 4.25. Predicted Gyrations for Projects with PG76-22 ..........................................145 Figure 4.26. Predicted Gyrations to Match In-Place Density for all Post-Construction Sampling Periods ...............................................................................146 Figure 4.27. Relationship between As-Constructed and Ultimate Density .....................148 xxi Figure 4.28. Two-Year Densification versus As-Constructed Density ...........................149 Figure 4.29. Comparison Between 3-2-2 Pine and Troxler Locking Point .....................162 Figure 4.30. Comparison of Average Pine and Troxler Density at 3-2-2 Locking Point...................................................................................................................163 Figure 4.31. 2-1 Locking Point Density versus As-Constructed Density.......................163 Figure 4.32. 3-2-2 Locking Point Density versus 2-Year Density ..................................164 1 CHAPTER 1 INTRODUCTION AND RESEARCH APPROACH 1.1 BACKGROUND The Superpave mix design system, a product of the Strategic Highway Research Program (SHRP), was released in 1994. The Superpave mix design system for hot mix asphalt (HMA) includes: binder specifications, aggregate property specifications, design gradation ranges, a laboratory compaction procedure, specifications for volumetric properties and an evaluation of moisture sensitivity. These specifications are to act in concert to provide a system of checks and balances to ensure the resulting HMA is durable and rut resistant. Durability would include such performance parameters as resistance to low temperature and age related cracking, resistance to raveling or other surface wear and resistance to moisture damage. Rut resistance refers to resistance to permanent deformation resulting from shear flow of the hot mix asphalt; permanent deformation or rutting of the subgrade due to insufficient pavement structure is not included. The Superpave Mix Design System was designed to account for differing traffic and environmental conditions. Central to the Superpave mix design system is the Superpave gyratory compactor (SGC). The SGC is used to compact trial HMA mixtures to a design number of gyrations in the laboratory in order to allow an evaluation of the volumetric properties of the compacted sample. The volumetric properties evaluated include: air voids, voids in mineral aggregate (VMA), voids filled with asphalt (VFA) and dust to effective binder 2 content. Two additional parameters are included to examine the rate of densification: density at an initial number of gyrations (Ninitial) and density at a maximum number of gyrations (Nmax). The laboratory design air content is supposed to be related to the ultimate field density of the HMA. Ultimately, the overall performance of an HMA pavement is highly dependent on the pavement structure and the construction quality. The pavement structure is evaluated in the pavement thickness design procedure, a separate topic. The ability to construct the HMA pavement layers should be, as much as possible, considered in the mix design procedure. The purpose of this research was to verify the relationship between laboratory testing and field performance with regards to the SGC, and, where needed, to provide alternative recommendations. 1.2 RESEARCH PROBLEM STATEMENT When the Superpave mix design system was initially released in 1994, it included 28 different design gyration (Ndesign) levels for the SGC, representing seven traffic levels for each of four climates (1). Traffic levels were represented by equivalent 18- kip single axle loads (ESAL) accumulated during a 20-year design life. Differing climates were represented by the average 7-day high air temperature for the project site. Ndesign increased as either design ESAL or high air temperature increased. In 1999, the Federal Highway Administration Superpave Mixture Expert Task Group recommended a consolidation of the original 28 Ndesign levels to 4 Ndesign levels (the author was present at this meeting). This consolidation was primarily based on research conducted in two studies (2,3). The consolidation eliminated differing 3 Ndesign levels for differing climates and reduced the design traffic to 5 ranges, 2 of which utilize the same Ndesign level. One of the studies did not address the magnitude of the Ndesign levels with respect to field performance, but rather differences in the gyration levels which resulted in significant differences in the resulting volumetric properties (2). The other study was based on the performance of a limited number of field sections (3). The American Association of State Highway and Transportation Officials (AASHTO) adopted the recommended changes to the SGC compaction procedure of the Superpave mix design procedure in 2000 (4). There is still concern that the current Ndesign levels do not maximize field performance. The optimum asphalt content for a given blend of materials is selected at 4 percent air voids, based on laboratory samples compacted to Ndesign, assuming the resulting mixture meets the other criteria of the Superpave mix design system. The asphalt content of HMA is critical to its performance, too much asphalt and the mixture is likely to suffer excessive permanent deformation. Too little asphalt and it maybe difficult to achieve field compaction; and the pavement may develop premature cracking, raveling and/or other distresses related to durability. The locking point concept has been proposed as an alternative to Ndesign. The locking point is believed to represent the point where the aggregate skeleton ?locks? together and further compaction results in aggregate degradation. 1.3 OBJECTIVE The three objectives of this research were 1) to evaluate the field densification of pavements designed using the Superpave mix design system, 2) to verify or determine the 4 Ndesign levels to optimize field performance, and 3) to evaluate the locking point concept. 1.4 SCOPE This study included a literature search and extensive laboratory and field testing. Samples were collected, tested and analyzed from 40 field projects at the time of construction. The projects were selected in a total of 16 states. The projects represent a wide range of traffic levels, binder grades, aggregate types, and gradations. Each project was visited at 5 time intervals after construction: 3 months, 6 months, one year, two years and four years. Coring and distress surveys were conducted at each evaluation interval. In total, approximately 4,085 SGC samples and 5,670 cores were tested. Data obtained from the SGC samples and field cores, as well as traffic data provided by the agencies were analyzed to provide recommendations for the Ndesign compaction levels and use of the locking point as an alternative to Ndesign. 5 CHAPTER 2 LITERATURE REVIEW Literature was reviewed for this study related to the history of HMA design, the densification of HMA pavements, gyratory compaction, Ndesign and the locking point concept. 2.1 A BRIEF HISTORY OF HMA MIX DESIGN PRIOR TO SUPERPAVE 2.1.1 Proprietary Mixes The first asphalt pavement constructed in the United States (U. S.) was built in Newark, New Jersey in 1870 (5, 6). This pavement was constructed with asphalt binder and rock asphalt imported from Europe (6). In 1876, President Grant appointed a commission of the U. S. Army Engineers to recommend paving materials for Washington, D. C (5). Based on this study, the first ?sheet asphalt? pavement was constructed later that same year on Pennsylvania Avenue using Trinidad Lake Asphalt, clean sand and mineral filler (6). Amzi Alonzo Barber purchased the rights to collect and remove Trinidad Lake Asphalt. Barber was awarded a portion of the Washington, D. C. paving contracts. In 1883, he formed the Barber Asphalt and Paving Company. E. B. Warren was one of the founders of the Barber Asphalt Company, which was engaged in the import of Trinidad Lake Asphalt. Captain Francis V. Greene was an Assistant 6 Engineer in charge of paving Washington, D. C. He later joined and became president of the Barber Paving Company. Barber-Greene was one of the early manufacturers of paving equipment (5). Two other paving companies were organized by members of the Warren family: Warren-Scharf Paving Company (1884) and The Warren Chemical and Manufacturing Company. In 1899, the Barber Asphalt Company, the Warren Chemical and Manufacturing Company and the Warren-Scharf Paving Company all merged. Hveem (5) referred to this group as the ?Asphalt Trust?. The remaining independent, National Asphalt Company, was brought into the group as the General Asphalt Company of America. Barber eventually withdrew from the trust to establish the A. L. Barber Company, which maneuvered to secure the rights to Bermudez Lake Asphalt, another natural asphalt source found in Venezuela. Until the beginning of the 20 th century, there is little evidence of design procedures or standardized tests. The asphalt ?trust? mainly produced sheet asphalt using fluxed Trinidad Lake Asphalt. In 1905, the first textbook on asphalt pavements was published by Clifford Richardson (5, 6). Mr. Richardson, a chemist by training, began his career with the U. S. Department of Agriculture. He then became engineer inspector for the District of Columbia and later was employed by the Barber Asphalt Paving Company (7). Richardson proposed the following specification for sheet asphalt (8): 1. Asphalt penetration of 30 to 90 (0.1 mm) at 78?F for the surface course and 20 units higher for the binder or leveling course. 2. The mixture consist of refined natural asphalt, fluxed to the above consistency, sand of an appropriate grading, and mineral filler such as 7 rock dust or Portland cement. In this case refinement refers to the removal of water and excess organic matter. 3. The sand has 100 percent passing the No. 10 screen, at least 15 percent passing the No. 80 sieve and at least 7 percent passing the No. 100 screen. The sand contains less than 1 percent clay. The sand is to be mixed with 9.5 to 12.0 percent asphalt. The penetration test was a recent invention, prior to which time asphalt consistency was evaluated by chewing. H. C. Bowen of the Barber Asphalt Paving Company invented the Bowen Penetration Machine in 1888. A. W. Dow, an inspector for the District of Columbia, designed another version of the penetrometer in 1903. Dow also invented the ductility test. Aggregate gradations, the penetration test for asphalt consistency and asphalt content determination by extraction using carbon disulfide made up the early asphalt tests (5, 8). The one test Richardson mentions to aid in the determination of optimum asphalt content is the Pat Test. The Pat Test consisted of a visual examination of a piece of Manila paper which had been pressed against a sample of HMA. A light stain indicated too little binder; a heavy stain indicated too much binder; and a medium stain indicated the optimum asphalt content (9). The first HMA, which incorporated coarse aggregate, originated in 1901 with a patent application by Frederick J. Warren for ?Bitulithic? pavement. A second patent was issued in 1903. Bithulithic pavements used tightly specified dense gradations with a maximum aggregate size of up to 3 inches. The large aggregate size tended to result in low asphalt contents, as compared to sheet asphalt. Also, the dense gradation allowed the use of softer asphalt cement resulting from the refinement of petroleum oil, mainly from 8 California, termed oil asphalt (5, 6, 10). A patent for ?Warrenite? pavement, which incorporated a thin layer of sheet asphalt laid on top of hot Bitulithic pavement soon followed (5, 9). The sheet asphalt tended to prevent the steel rimmed wheels of the day from fracturing the large coarse aggregate particles found in the Bitulithic pavement, and allowing water to enter the pavement. Since the sheet asphalt was placed in a thin layer, it was not as prone to rutting as pavements constructed solely of sheet asphalt. The City of Topeka, Kansas developed a mix consisting of sheet asphalt with a limited amount of ? inch coarse aggregate added in an attempt to avoid paying royalties on the Warren Brothers patents. This mix became known as the ?Topeka? mix. In 1912, The Warren Brothers filed suit against the City of Topeka for patent infringement. The federal court in Topeka, Kansas ruled that it was possible to construct an asphalt pavement that did not infringe on the Warren Brother?s patents if the nominal maximum aggregate size was less than ? inch (5, 6, 10). Davis (10) credits this ruling for the predominance of small (less that ? inch) top size aggregate surface mixes used today. From 1900 until the early 1920?s the majority of the asphalt pavements constructed were constructed with one form or another of proprietary HMA. Davis (10) notes, that there was little incentive for the companies, such as the Warren Brothers, to explain their design procedures. From 1920 until 1940, the use of HMA pavements continued to grow. During this period pavements were typically designed with one of four techniques (6): 1. Sheet asphalt produced by Richardson?s or similar procedures, 2. Bitulithic, Warrenite or one of the other HMA mixes patented or trademarked by the Warren Brothers, 9 3. The Skidmore method which was similar to the Warren Brother?s mixes, but had the addition of mineral filler to fill voids, or 4. The Hubbard-Field Method developed by Prevost Hubbard and Frederick Field (described below). 2.1.2 Hubbard-Field Prevost Hubbard and Frederick Field developed a mix design method for the fine fraction (100 percent passing the No. 10 screen) of sheet asphalt and sand base mixes. The maximum load required to force a 2 inch diameter by 1 inch tall compacted sample through a 1.75-inch diameter orifice was plotted as a function of asphalt content. The maximum load was termed a ?stability? value. The method was reportedly still in use by several states in the 1970s (5, 6, 9, 11). From the late 1930?s through approximately 1960, the modern philosophies of HMA mix design were developed, including: Hveem, Marshall, Texas Gyratory, and Corp of Engineers Gyratory Testing Machine. 2.1.3 Hveem Method Francis N. Hveem was first exposed to asphalt as a young employee of the California Division of Highway. In 1927 he oversaw his first oil-mix job. Oil-mixes were road oil, slow curing cutback asphalt, mixed with gravel using a grader and rolled. Shortly thereafter, Hveem transferred to the Central Laboratory in Sacremento, California. By 1929, Hveem observed that coarser gradations tended to require less road oil than finer gradations and made the connection that the surface area of the aggregate 10 varied with gradation. Hveem identified a method for calculating (estimating) the surface area of aggregate developed by a Canadian engineer, Captain L. N. Edwards for Portland cement concrete mixes (5, 12). Hveem realized that in addition to surface area, the optimum asphalt content, or at least the point where the optimum asphalt content was exceeded and stability decreased was affected by the surface texture of the aggregate. A ?surface factor? was used by Hveem in combination with the calculated surface area to determine the optimum asphalt content. Although an experienced engineer could adjust for texture and absorption of various aggregates, Hveem later developed the centrifuge kerosene equivalent (CKE) test to estimate the surface constant (a combination of surface area, absorption and adjustment for surface texture) of the fine aggregate. A 100 g sample of the fine aggregate (100 percent passing the No. 4 sieve) was saturated in kerosene. The sample was then subjected to 400 times gravity in a centrifuge (13) [later this was reduced to 200 times gravity (11)], after which the aggregate was weighed to determine the percent of kerosene retained by mass of dry aggregate. If the fine aggregate type was similar to the coarse aggregate, then the bitumen index or the quantity of asphalt required to coat one unit of the area of aggregate could be determined directly from the CKE test; otherwise a separate test could be performed to determine the surface factor of the coarse aggregate (13). The coarse aggregate absorption test was performed by soaking a sample of the coarse aggregate in S. A. E. 10 oil for five minutes, and then allowing the sample to drain for 15 minutes at 140?F before determining the percent of retained oil. The coarse aggregate surface factor was used to correct the fine aggregate surface factor. These procedures, either the surface area calculation or the surface factors could be used to estimate optimum binder content. Correction factors were also included for aggregate specific gravity and the viscosity of the asphalt. Hveem did observe that a smaller film thickness of asphalt was required for smaller particles than for larger particles. Hveem stated that the CKE method indicated the optimum asphalt content in 95 percent of cases (5, 13). Hveem also wanted to evaluate the stability of the HMA. He hypothesized that depending on the roughness and angularity of the aggregate, the film thickness at which the particles would become overly lubricated by the asphalt and therefore unstable would vary (13). Hveem was not satisfied with the Hubbard-Field method in use at that time. This led to the development of the first Hveem stabilometer in 1930. The stabilometer evolved into a hydraulic device into which a compacted sample of asphalt was loaded. The sample was loaded vertically on its flat surface and the radial force transmitted to the surrounding hydraulic cell is measured. The stability value is calculated according to Equation 1: 222.0 ) 2.22 = ( v P 2 + ? h h P DP S (1) v h 2 = displacem where, P = vertical pressure (400 psi), P = horizontal pressure at a vertical pressure of 400 psi, and D ent of sample in number of turns of handle. The use of the stabilometer required a compacted sample 4 inches in diameter and 2.5 inches tall. Initially an impact compaction method, consisting of an 8-lb hammer dropped 5 inches which applied blows to a 2-inch diameter tamper around the perimeter of the mold, was used. Vallerga and Lovering (12) state, ?This method was used for 11 12 o turn e) would realign aggregate particles in a similar manner to a e 2) quote Hveem?s own summary of his mix design philoso th . ousness to water, a high asphalt content, broadly speaking, the richer the th a sufficient quantity of fine several years, but when cores were cut from the pavement and the Stabilometer value compared with specimens of the same material compacted in the laboratory, it was found that the laboratory specimens invariably had a considerably higher stability.? This led t the development of the kneading compactor which pneumatically loads a tamping foot with a cross section of one quarter of the mold area while rotating the mold 1/6 of a between each tamp. It was felt that the ?kneading action produced by the foot (not covering the entire surfac rubber tire roller or car. The optimum asphalt content by the Hveem method was determined using a pyramid scheme. First, the asphalt contents for which moderate to heavy bleeding wer observed on the surface of the compacted sample were eliminated. Next, any asphalt contents that failed the minimum stability value were eliminated. Finally, the highest asphalt content that had at least 4 percent air voids was selected as the optimum (11). Vallerga and Lovering (1 phy in 1937 as follows, ?For the best stability, a harsh, crushed stone with some gradation, mixed wi only sufficient asphalt to permit high compaction with the means available For greatest resistance to abrasion, raveling, aging and deterioration, and impervi better. For impermeability, a uniformly graded mixture wi sand (fine sand is more important than filler dust). 13 the rule to use a dense, uniformly graded mixture without an excess of sing [Currently, we would describe ?uniformly? graded as ?well? or ?dense? graded]. Graphically, this philosophy is summarized in Figure 2.1. For non-skid surfaces, a large quantity of the maximum sized aggregate within size limits used. For workability and freedom from segregation, a uniformly graded aggregate. To reduce the above factors to as simple a consideration as possible, it seems to be the best dust and to add as much oil or asphalt as the mixture will tolerate without lo stability.? Figure 2.1. Stability and Durability as a Function of Asphalt Content (12). 2.1.4 Marshall Mix Design Bruce G. Marshall began the development of what later became known as the Marshall mix design procedure around 1939 while employed by the Mississippi State 14 rps s. The initial compaction effort was 15 blows esign approximately 100 psi. By the end of World War II, raft bard-Field method as well as a ethod n Highway Department (11). Marshall developed the stability test; flow measurements were added by the U. S. Army Corps of Engineers. Marshall was retained by the Co during their studies (6). Initially, samples of HMA for the stability and flow tests were compacted with a modified American Association of Highway Officials (AASHO), California Bearing Ratio (CBR) field hammer. The modified AASHO hammer consisted of a 10 pound hammer (weight) dropped 18 inches; the load was transferred to the sample through a 1.95-inch diameter foot. Samples were compacted in a 4-inch diameter mold with a target compacted height of 2.5 inche of the modified AASHO distributed across one face of the sample followed by a 5000 pound static load held for 2 minutes (14). The Corps of Engineers was charged with selecting a method of HMA mix d to deal with the increasing tire pressures found on military aircraft. Aircraft weights began increasing during World War II. As the weight of the aircraft increased, tire pressures were also increased to minimize the size of the landing gear. At the beginning of World War II, tire pressures were tire pressures had increased to approximately 200 psi. Currently, some military airc have tire pressures of 350 psi (15). In a previous study, the Tulsa District of the U.S. Army Corps of Engineers recommended the Hubbard-Field method of HMA mix design. In 1943, the Waterways Experiment Station was charged with evaluating the Hub m utilizing the field CBR hammer (14). At this time the Marshall method had bee used by some southern states for up to four years (15). 15 ve stability test; further, the Hubbard-Field test was not read more portable. e ethod was selected for additional study to evaluate the following objectives (14): 1. For both sand asphalt and HMA evaluate the effect on test properties from: ation grade of asphalt cement 2. ere is a correlation between laboratory compaction and field - ith asphalt. In addition le, 37,000 In the first phase of the study begun in 1943 (14), comparisons were performed between the Hubbard-Field and Marshall mix design methods using a wide range of asphalt materials. From this study it was concluded that the Marshall Stability test ga comparable results to the Hubbard-Field ily adaptable to the field CBR equipment; and the Marshall apparatus was also Th refore, the Marshall m a. Aggregate gradation b. Type of filler c. Mixing temperature d. Penetr e. Compactive effort. Determine if th compaction. 3. Determine the relationship between the Marshall method and the Hubbard Field method. The Marshall test properties selected for evaluation included stability and flow, total unit weight, aggregate unit weight, percent voids total mix, percent voids aggregate only (essentially voids in mineral aggregate) and percent voids filled w to evaluating asphalt mix design properties, the Corps were also charged with evaluating the required pavement thickness for three different wheel loads, 15,000 lb sing lb single and 60,000 lb double on differing subgrade types. 16 ed ouble surface treatment. HMA sections utilized from both size ced with h ach Test sections were constructed to allow the laboratory properties to be compar with field performance. The test tracks were divided into 8 major sections to accommodate three mix types and three subgrade qualities. The three mix types were HMA, sand asphalt and d crushed limestone and uncrushed gravel coarse aggregate with a maximum particle of ? inch. Siliceous sand from a river pit and from a Mississippi river sand bar were used for fine aggregate. Three subgrade materials were used in the study: crushed limestone (high quality), sand-loess (medium quality) and sand-clay-loess (low quality) were used for the evaluation of the minimum required pavement thickness. Only the HMA produ crushed limestone was placed on all three subgrade materials; the HMA produced wit uncrushed gravel was only placed on the high quality crushed limestone subgrade. E of the 8 sections, except the two double surface treatment sections, was further subdivided into three thicknesses, each 90 feet long. The total pavement thicknesses were 1 ?, 3, and 5 inches for the HMA and 2, 4, and 6 inches for the sand asphalt. To evaluate the effect of filler on Marshall stability, each pavement thickness section was further subdivided into three 30 foot sections with three different levels of limestone mineral filler addition to the HMA or sand asphalt: none, some and high. Finally, at each level of mineral filler content, the HMA or sand asphalt was produced at three asphalt contents: that which produced the maximum stability using the previously described compaction procedure, and 10 and 20 percent below optimum. Previous experience with a test section in Marietta, Georgia indicated that the optimum asphalt content determined from the maximum stability value would be too rich (high in asphalt), 17 re 10 feet long. All of the main sections were produced with a 120- than a separate lane for each wheel load. It is interesting to note that the lanes r o loads. The net tire contact pressures were 106, 146 and 139 psi for the 15,000, 37,000, and 60,000-lb wheel loads, respectively. Net pressures leading to too low of in-place air voids under traffic. The sections for the different asphalt contents we 150 pen binder. By today?s specifications, this is a very soft binder, probably softer a PG 58-28. Additional studies, including the use of gap gradations were conducted in the turnarounds. In total, the two straightaway sections were 850 feet long and 60 feet wide, allowing for were paved perpendicular to the direction of traffic. The ten foot width of the paving lane, which was 60 feet long, became the ten foot length of the test lane for a given wheel load. Traffic loads were applied using a Model C Tournapull, essentially the engine and drive wheels of a modern scraper or pan. A 12-cubic yard scraper was loaded to provide 15,000 lbs load on each of its two wheels. This setup was used to provide 3500 coverages across an approximately 12-foot lane width with the 15,000 lb wheel load. A specially built cart was built to apply the 37,000 and 60,000-lb wheel loads. A single (fo 37,000-lb load) or dual (for the 60,000-lb load) 56-in diameter wheel was mounted in the center of the cart (Figure 2.2). The load was applied to a 4-foot or 6-foot lane width for the 37,000 lb or 60,000 lb load, respectively. The cart had two additional wheels which were loaded to 10,000 lb each, but these as well as the Tournpull drive wheels (loaded t 14,000 lbs) tracked outside the test lanes. A total of 1500 coverages were applied with the 37,000 and 60,000-lb wheel 18 es were used to account for the block nature of the tire tread. The majority of the coverag were applied in warm weather. Figure 2.2. Model C Tournpull w rafficking by visual observations and coring. rutting hness, upheaval and longitudinal mo and pronounced. The 4-inch diam flow. to this current study (14): 1. The test property relationships developed during construction and subsequent trafficking were similar to those developed from laboratory compaction. 2. There was an indication that the number of roller passes required to match the laboratory density varied with the mix type and asphalt content. ith Specially Built Loading Cart (14). The performance of the test sections was monitored throughout t Visual observations included: tire printing (bleeding), and shoving, cracking, settlement, roug vement. Four levels were used to quantify the observations: none, faint, well-defined eter cores were tested for density and stability and The following is a summary of the conclusions from the Corps study which relate 19 3. ate gradation was believed to be of lesser importance than other factors in the design of good performing HMA. 4. In all cases, density increased with the application of wheel passes (Figure 2.3). Density increased rapidly at first, and then more slowly after the first few hundred passes. Regardless of initial, as-constructed, density, the densities of identical mixes subjected to three different wheel loads were nearly identical after 1500 passes. Aggreg Circle = Optimum ? 20% Triangle = Optimum -10% Square = Optimum Fig Medium Filler Content (14). 6. ure 2.3. Traffic Compaction Data for Mix 11, Crushed Limestone with 5. The range of asphalt content that produces satisfactory performance is approximately ? 1.0 percent. The optimum asphalt content selected at 4 percent air voids and 80 percent VFA for HMA (6 percent air voids and 70 percent VFA for sand asphalt) was in reasonable agreement with those deemed acceptable based on the field test sections, but on the low end of the range. 20 tatic load held for 2 minutes, as well as a modified ere conducted to examine other compaction efforts that might phalt content from the remaining four parameters were averaged to determine re 7. The as-constructed density was approximately equivalent to the density obtained in the laboratory from the original compaction effort, 15 blows to a 1.95-inch diameter foot plus a 5,000 lb s compaction effort, 15 blows on each face with a 10-lb hammer falling 18 inches with a 3 7/8-inch diameter foot. This density was approximately 2 percent less than that obtained with 50 blows on each face with the modified compaction effort. 8. Tire pressure is more important than wheel-load in its effect on the performance of the pavement. No difference in performance was noted for net tire pressures ranging from 106 to 146 psi. Additional studies w account for the densification which occurred under traffic. From this effort, the familiar compaction effort, 50 blows to each face with a 12.5-lb hammer falling on a 3 7/8-inch diameter foot, was developed. This was later changed back to a 10 lb hammer. Five properties were selected for design: stability, flow, unit weight, air voids and VFA. Flow was only used as an evaluation of the plasticity of the mix (maximum value of 20). The optimum as the design asphalt content. In summary the Corps of Engineers (14) note, ?The results of this study indicate that the quantity of asphalt is the most important factor in a paving mixture. Where the is too much asphalt in the mix the resultant pavement will ?flush? and the pavement will rut and shove under traffic. Too little asphalt produces a brittle pavement that will crack 21 lt as e in si t and , ific res d ture.? and ravel. From the standpoint of durability, it is desirable to include as much aspha possible.? As mentioned previously, aircraft tire inflation pressures continued to increas the late 1940?s and early 1950?s. Tire pressures doubled from the approximately 100 p net tire pressure used in the first field study to 200 psi. White reports (15), additional tests were conducted on the original test sections using both 30,000 lb wheel load with a 200 psi tire pressure and 15,000 lb wheel load with a 240 psi tire pressure. From these efforts it was determined that 69 blows from a 10-lb hammer falling 18 inches on a 3 7/8- inch diameter foot were appropriate for the increased tire pressures. This was later adjusted to the 75-blow Marshall. McLeod (16) first suggested the concept of designing for minimum VMA to ensure durability in 1956. VMA is the total void space filled with either air or asphalt between the compacted mineral aggregate, which is believed to be related to durability. He argued that VMA and VFA should be calculated with the effective binder conten aggregate bulk specific gravity to avoid errors with absorptive aggregates (16). In 1957 McLeod reaffirmed his belief that the effective binder content and aggregate bulk spec gravity should be used to calculate the VMA and air voids of the compacted HMA sample (17). McLeod stated: ?Values for percent voids in mineral aggregate and for percent air voids can be defined precisely for compacted bituminous paving mixtures that are made with non-absorptive aggregates.? He added: ?For compacted paving mixtu that contain absorptive aggregates, values for percent voids in the mineral aggregate an for percent air voids, should be calculated by means of (a) the ASTM bulk specific gravity of the aggregate, and (b) the effective bitumen content of the paving mix 22 of f the Marshall mix design ually, mechanical Marshall Hammers were developed to reduce the effort require effort th during procedu nch diamete using th esign HMA. Leahy and McGennis (6) provide a rare quote of Marshall?s own mix design of on No limits can be established for VMA, for universal s materials to many types and gradations of aggregates.? McLeod?s objections to the use of apparent and effective aggregate specific gravities (which are substantially easier to measure) result from their failure to differentiate between the portion of the binder that is coating the aggregate particle and the portion the binder that is absorbed in the aggregate. Without this differentiation, it is difficult to relate observations from the laboratory design to field performance in terms of both permanent deformation and durability. In 1962, the Asphalt Institute published a new version of MS-2 that included the first ?modern? version o procedure including volumetric analysis based on effective binder content (18). Event d by the operator to produce samples. These tended to produce less compactive an a hand-held hammer. This is attributed to the operator moving the handle compaction, producing a slight kneading action (19). The Marshall mix design re was expanded to include 1 ? inch maximum aggregate by developing a 6-i r mold with a 75-blow compaction effort (20). By 1984, 38 out of 50 states were e Marshall mix design procedure to d philosophy: ?The ultimate result in the improvement of aggregate gradation is the reduction the VMA. VMA should be reduced to the lowest practical degree. This reducti results in a superior pavement structure as well as to reduce the quantity of asphalt required in the mixture. application, because of the versatile application of bituminou 23 2.1.5 T I design a to develop a means of r the laborato the uld ate the degradation that occurs in the field. m that as two 24-inch handles attached at a 75-degree angle to one exas Gyratory Method n 1939, the Texas Highway Department initiated a research program into the nd field control of HMA (21). The first goal of the research was compacting samples in the laboratory. The following criteria were listed fo ry compaction method: 1. Method must be adaptable to field control of HMA mixes. 2. The method should yield essentially the same density that is obtained in finished pavement. Since pavements continue to densify under traffic, the laboratory density should approximately match the ?ultimate? density after some time on the road, ?and is the goal of any compaction method.? 3. The aggregate breakdown that occurs during laboratory compaction sho approxim A number of compaction devices were evaluated. These methods applied shear to the surface of the sample. It was desirable to develop a method that applies shear throughout the sample while holding the faces of the sample, to which compressive forces are applied, parallel. The Texas Gyratory Molding Machine was developed fro this effort. Using this device, Ortolani and Sandberg (21) state, ?The aggregate is oriented into its most dense position by applying specimen shear at low initial pressures.? The original Texas Gyratory Molding Machine consists of two loading heads are held parallel to one another. The lower loading head is connected to a 30 ton jack. The molding cylinder h 24 another (Figure 2.4). The handles are used to manually impart the gyratory action; a 50-lb compressive til ing per guide ring limits the mold?s vertical movement to ? inch. First a load is applied to the sample; then the handles are used to impart a gyratory action un 3 revolutions were completed. This is to be repeated until movement of the mold cylinder is extremely difficult. At this point, one stroke of the jack handle should increase the gauge pressure to 100 lbs. This indicates the sample has reached the pro degree of compaction. Figure 2.4. Manual Texas Gyratory Molding Machine (21). In 1945, the Texas Highway Department took over 400 cores from around the state from pavements which were 1 to 12 years old in order to compare in-place pavement densities to those determined using the Texas Gyratory Molding Machine. In-place densities at the time of construction were also available; these averaged 3.8 percent less than the density of the samples compacted in the Texas Gyratory Mold Machine. The cores which were taken after 1 to 12 years of traffic averaged 0.8 percent less than the laboratory samples. There was variability in the data. One coarse-graded ing 25 ep in the pavement tructure, was 2.3 percent less than the laboratory compacted samples (21, 22). ter automated. In 1974, the method s ore times. This to e channelized high-pressure tire traffic. The goals of this research were to develop a compactor that could simulate in-place pavement density after traffic as well as produce laboratory samples with Marshall Stabilities similar to those obtained from cores. pavement?s density was 3.3 percent less than the laboratory compacted samples after one year of traffic. Another base layer, approximately 3 inches de s The Texas Gyratory Molding Machine was la was adopted as ASTM D 4013, ?Standard Test Method for Preparation of Test Specimens of Bituminous Mixtures by Means of Gyratory Shear Compactor (23).? When using the Texas Gyratory Compactor, the number of gyrations is variable in groups of three gyrations applied at one gyration per second. First, a 50 psi vertical pressure, termed the gyration pressure, is applied to the sample. Next, the sample i gyrated three times at an angle of 6 degrees. At this point if one stroke of the hydraulic pump increases the vertical pressure to 150 psi, the gyrations are complete. Otherwise, the pressure is reduced to 50 psi and the sample is gyrated three m process is repeated until one stroke of the hydraulic pump causes the vertical pressure increase to 150 psi. Finally, the vertical pressure is increased to 2500 psi at the rate of one stroke per minute. This is termed the end pressure. Once 2500 psi is reached, th pressure is immediately released and the sample extruded (24). 2.1.6 Corps of Engineers Gyratory Compactor McRae (25) presented the development of the Corps of Engineers Gyratory Compactor to simulate the in-place pavement densification which occurred under 26 Stabilities of samples compacted with the Marshall hammer tended to be higher than the stabilities of pavement cores of the same mixture tested at the same density. This was believed to be related to differences in the aggregate orientation. The Corps of Engineers Gyratory Compactor was based on the Texas Gyratory Molding Machine, discussed previously. The gyratory action is provided mechanically by a pair of rollers riding on a flange connected to a sleeve surrounding the samples mold (Figure 2.5). The arm, to which the two rollers are affixed, is rotated by an electric motor. The initial angle of gyration can be adjusted using a thumb screw attached to the lower roller. The pressure of the upper roller is adjustable using an air over oil chamber. A hydraulic jack is used to provide a variable vertical pressure, up to 300 psi, on the sample. The combined action produces a ?fixed-deformation variable Figure 2.5. Schematic of Compaction Head for Corps of Engineers Gyratory Compactor (25). 27 ls included a heated jacket around the sample mold. Figure 2.6 shows a comparison between the densities of samples compacted with varyin ngineers even 150-blow Marshall samples; howev stress? type compaction. The sample is compacted at a rate of five gyrations per minute. Later mode g laboratory compaction efforts with both the Marshall Hammer and Corps of Engineers Gyratory Compactor and field densities after varying levels of accelerated loading. The author notes that the as-constructed density was approximated by both the 50-blow Marshall and 5 gyrations with a 100 psi vertical load of the Corps of E Gyratory Compactor (left side of Figure 2.6). The author also notes that the in-place pavement density after 2615 coverages exceeded er, the in-place density could be exceeded by 60 gyrations at either 200 or 300 psi. It was also noted the Marshall stabilities of samples produced with the Corps of Engineers Gyratory Compactor more closely approximated those of field samples (right side of Figure 2.6). Figure 2.6. Comparison of Laboratory and Field Density and Stability Values (25). 28 lt for ensity versus asphalt content can be used to determine the t which the mix becomes plastic. As the compaction effort increases, the asphalt content at which the mix becomes plastic decreases. This is graphically illustrated in Figure 2.7. The ratio of the stress on the upper oil roller versus the vertical stress might be another indicator of mix stability. In 1958, McRae and McDaniel (26), reported on additional advancements with the Corps of Engineers Gyratory Compactor. Rate of gyrations was studied and observed to have little effect on sample density. The machine was modified to record the gyratory motion of the sample during compaction. Initially, the angle of gyration would decrease from the level set prior to beginning the test; indicating densification of the mix. This densification would be a combination of that which occurs at the time of compaction and that which occurs under traffic. The pressure in the oil roller would increase during this phase. When a critical density was achieved, the specimen would become plastic and the t the number of gyrations before this occurred could be related to traffic. ples The author goes on to outline a framework for selecting the optimum aspha HMA. A plot of aggregate d asphalt content a angle of gyration would again increase and the oil-roller pressure would drop. It was elieved thab Recommendations were also developed to prepare samples with similar densities to samples compacted with the Marshall Hammer: 50-blows was approximately equivalent to samples compacted in the gyratory with a 100 psi vertical pressure and 1 degree initial angle compacted to 30 gyrations and 75-blows was approximately equivalent to sam compacted in the gyratory with a 200 psi vertical pressure and 1 degree initial angle compacted to 30 gyrations. Figure 2.7. Aggregate Density as a Function of Asphalt Content and Compaction Level (25). ed for a ?variable stress and variable shear strain testing capability? (27). ure framework of the French mix design procedure for the Texas Gyratory Molding Machine and the GTM, the The Corps of Engineers Gyratory Compactor was later renamed the Corps of Engineers Gyratory Testing Machine (GTM) and adopted in 1974 as an ASTM D 3387, ?Standard Test Method for Compaction and Shear Properties of Bituminous Mixtures by Means of the U. S. Corps of Engineers Gyratory Testing Machine (GTM) (23)?. Additional research led to the development of an air roller to replace the oil roller which allow 2.1.7 French Design Proced Bonnot (28) outlined the HMA. The French use their Gyratory Shear Compacting Press (PCG) to evaluate the workability of HMA. Similar to 29 30 ld parallel during compaction with the mold forming an fixed and the other describes a cone as shown of te. ends of the HMA sample are he oblique cylinder. One end of the sample is in Figure 2.8. The sample is compacted in a 160 mm diameter mold with a final sample height of approximately 150 mm. During compaction, a vertical compressive pressure 0.6 MPa (87 psi) is applied to the sample and the angle of gyration is fixed at 1 degree from vertical. The sample height and the force required to maintain the 1 degree gyratory angle are recorded with each gyration. Assuming a fixed sample mass and mold diameter, the density of the sample can be estimated at each gyration. Samples are generally compacted to 200 gyrations at a rate of 6 gyrations per minu Figure 2.8. Compaction Principle of the PCG (28). Correlations studies were conducted between the density obtained with the PCG and the in-place density achieved with a rubber tired roller at a given layer thickness. Equation 2 was developed for comparing the field compaction for lifts ranging in thickness from 3 to 12 cm to an equivalent number of gyrations in the PCG. (2) re, pg NekN ??= whe 31 g = number of PCG gyrations, r 10 ton vibratory a given al maximum density) varies of . Additional performance ot N k = factor for compactor type; 0.0625 for rubber tired rollers and 0.25 fo rollers operating at 25 to 30 Hz, e = layer thickness, (mm), and N p = number of rubber tired roller passes. Using this equation, it is possible to estimate the obtainable in-place density using compaction effort. For instance, the achievable density of a 38 mm thick surface mix using 8 passes of a vibratory roller would be estimated at 76 gyrations of the PCG. The target in-place air voids (air voids = 100 ? percent of theoretic with climate, it is lower (3 to 4 percent air voids) for a cold mountainous region than it is for a hot region (6 to 7 percent air voids). If the air voids at the calculated number gyrations is too high, the mix is unworkable and may be adjusted by: ? Increasing asphalt content, ? Increasing filler content, ? Substituting rounded fine aggregate, or ? Other gradation changes such as gap grading. If the air voids are too low, the mix could be made stiffer by doing the opposite. The PCG is used to develop the initial job mix formula testing is conducted depending on the application and may include: resistance to permanent deformation, predicted fatigue life, and resistance to moisture damage. Depending on the design conditions, these tests may be used to modify the design or simply verify minimum performance. Samples for performance testing are produced n 32 s nsity of laboratory-compacted ecim ? operties tation with the PCG but with a compactor using a laboratory scale rubber tired roller. Sample may be sawed or cored from the resulting slab. 2.2 SUPERPAVE GYRATORY COMPACTOR 2.2.1 Selection of the SGC for the Superpave Mix Design System One of the tasks faced by the SHRP researchers during the development of the Superpave Mix Design System was the selection of a laboratory compaction procedure. In the introduction to the selection process, Cominsky et al. (29) note, ?compaction is considered the single most important factor affecting the performance of asphalt pavements. Hughes (30) stated, ?It is important that the de sp ens approximate that obtained in the field in terms of (a) the structure of the mix and (b) the quantity, size, and distribution of the air voids.? Consuegra et al. (31) conducted a study on laboratory versus field compaction as part of the NCHRP project on the development of the Asphalt Aggregate Mixture Analysis System (AAMAS). Consuegra et al. (31) describe a major objective of their study to, ?ensure that laboratory mixtures will be fabricated in a manner that adequately simulates field compaction and, consequently, will yield reliable engineering properties. Thus, two goals emerged, matching field air voids and matching the engineering properties of field compacted samples. [This author notes that the engineering pr of laboratory compacted samples are probably influenced by both aggregate orien and the degree of aggregate degradation during compaction]. 33 s methods were discussed previously. The Arizona mples with a rapid impact load (1,200 cycles minute) and low contact pressure with the sample tilted at a slight angle (1 degree m vertical) to the applied load. The mobile steel wheel simulator used in this study tration (FHWA). It consisted of curved le. The curved foot consisted of a segment of a ch ld The research on the AAMAS system was completed in 1991, three years prior to the completion of the Superpave mix design system (32). The AAMAS research wa linked to the SHRP research to develop the Superpave system. AAMAS included a study to select a laboratory compaction procedure by Consuegra et al. (31). Loose mix was sampled from five projects, one each in Colorado, Michigan, Texas, Virginia, and Wyoming and approximately 25 field cores were taken from each project immediately after construction. Five laboratory compaction devices were used in the study: mechanical Marshall Hammer, California Kneading Compactor, Arizona vibratory- kneading compactor, Texas Motorized Gyratory Shear Type Compactor and mobile steel wheel simulator. Three of these vibratory kneading compactor compacted sa per fro was obtained from the Federal Highway Adminis foot that applied a static load to the samp circle, simulating the action of a steel wheel static roller. The laboratory compactive efforts with the five devices were varied to achieve the average in-place density determined for each of the field projects. The required compactive effort for the Marshall Hammer varied from 20 to 47 blows per face to mat the in-place air voids. Initially, the researchers planned to reduce the number of gyrations with the Texas Gyratory shear Compactor; however three gyrations, the minimum that can be used with the Texas Gyratory, resulted in lower air void contents than the field cores. Therefore, the gyration pressure and end pressure were varied to match the fie 34 e n d king aboratory and field air voids after two years of traffic, and aggregate orientation. Both the Marshall Hammer and the Texas Gyratory air voids. The gyration pressure was varied from 25 to 100 psi; 50 psi is the Texas standard. The end pressure was varied from 0 to 2500 psi; 2500 psi is the Texas standard. The Texas project required the least and the Virginia project the most compaction effort to match the field in-place air voids at the time of construction. The engineering properties of the pavement cores and laboratory samples wer evaluated by means of indirect tensile strength at 41, 77, and 104 ?F, repeated load indirect resilient modulus, and indirect tensile creep. The average differences and mea square error (MSE) between the test results on field cores and laboratory compacted samples were used to assess the best compaction method. MSE equally weights the variance of the test results and the square of the bias of the test results between the field and lab compacted samples. Based on these analyses, no single compaction metho always provided the best match with the test results for the field cores; however, the Texas Gyratory Shear Compactor was consistently better. The following lists the ran of the compaction devices (31): 1. Texas Gyratory Shear Compactor, 2. California Kneading Compactor, 3. Mobile steel wheel simulator, 4. Arizona vibratory-kneading compactor, 5. Marshall Mechanical Hammer. In addition to the evaluation of the engineering properties of samples produced using various compaction methods as compared to field cores, Von Quintus et al. (1991) present comparisons on compactability, l 35 ared to the d n wo target air void contents (4 and 11.5 low produced the same compactability rankings as observed in the field. Based on MSE, the California Kneading Compactor best matched the field air voids after two years followed by the Marshall Hammer, Texas Gyratory Shear Compactor, Arizona vibratory-kneading compactor and mobile steel wheel simulator. The mobile steel wheel simulator and Texas Gyratory Shear Compactor best simulated aggregate orientation as comp field cores. Based on these results and limited testing with the GTM, the AAMAS researchers (32) recommended either the Texas Gyratory Shear Compactor or the GTM for producing laboratory compacted samples for design and performance testing. The SHRP A-003A Contractor, the University of California at Berkley (33), conducted a study of the effects of laboratory compaction procedure on the rutting an fatigue properties of HMA. Three compactors were evaluated in the study: the Texas Gyratory Compactor, California Kneading Compactor and rolling wheel compactor. In addition, limited testing was conducted with the Corps of Engineers GTM and the Exxo Rolling-Wheel Compactor. Sixteen HMA combinations were evaluated in the study: two asphalt sources (same grade), two aggregate types (granite and chert), two asphalt contents (optimum based on California Kneading Compactor and optimum plus either 0.5 percent [granite] or 0.7 percent [chert]), and t percent). The optimum plus asphalt contents approximate that obtained from a 75-b Marshall design. Two primary tests were performed to evaluate the effect on rutting: static creep and shear creep; both tests were performed at two temperatures (40 and 60 ?C) and two stress levels (varied). Beam fatigue tests were performed on samples prepared using the California Kneading Compactor and the rolling wheel compactor. Since beam samples cannot be prepared with the Texas Gyratory Compactor, diametral 36 d istant fact e act. with the ly more pes d produce samples with very different engineering properties. fatigue tests were also performed using samples compacted with all three compaction methods. Fatigue tests were conducted in constant stress mode at two stress levels an two temperatures (0 or 4 ?C and 20 ?C). The California Kneading Compactor consistently produced the most rut-res samples and the Texas Gyratory the least rut-resistant samples. Dynamic modulus testing indicated that samples compacted with the California Kneading Compactor were in stiffer than samples compacted with the Texas Gyratory Compactor. This agreed with the findings from the AAMAS study (29). All three devices ranked all of the experimental variables in the same order, e.g., the granite aggregate was more rut resistant than the chert aggregate was. The California Kneading Compactor was more sensitive to aggregate type (angularity), than the Texas Gyratory Compactor was. Th greater rut resistance of samples compacted with the California Kneading Compactor was believed to be related to the development of greater aggregate inter-particle cont The Texas Gyratory Compactor consistently produced samples which had longer fatigue lives than those samples compacted in the California Kneading Compactor; the rolling wheel compactor samples produced an intermediate ranking between the two. The ranking of the experimental variables were different for samples compacted three different compactors. The Texas Gyratory Compactor was believed to be more sensitive to asphalt type than the California Kneading Compactor, but only slight sensitive than the rolling wheel compactor. Limited comparisons were performed with field cores from two projects in California. Testing with the Corps of Engineers GTM indicated that two different ty of gyratory compactors coul 37 ample P chers r d n ing Wheel ompa s s, and S s produced with the two different rolling wheel compactors were similar. SHR A-003A researchers (33) recommended the rolling wheel compactor. The resear emphasized the importance of having a single compaction procedure. This author believes that their decision was partially based on their desire to have a compaction procedure which could produce flexural beam fatigue samples. This study was later criticized for not having been correlated to field performance (29, 34). Based on the results from the AAMAS and SHRP A-003A studies, SHRP commissioned a third study which was conducted by Texas A&M University, the SHRP A-001 contractor (29). Five pavement sites were selected from the SHRP Special Pavement Studies (SPS)-5 and SPS-6 field tests. Approximately 30, 4-inch diamete cores were taken from each section. The average in-place air voids at the five sites varie from 3 to 8 percent, with a variation at each site of 2 to 5 percent. Four laboratory compaction devices were chosen for evaluation: the Texas Gyratory Compactor, Exxo Rolling Wheel Compactor, mechanical Marshall Hammer, and Elf Linear Kneading Compactor. The complete matrix of tests for all sites were only performed with samples compacted using the Texas Gyratory Compactor and the Exxon Roll C ctor. The laboratory compacted samples were produced with laboratory prepared HMA. Laboratory compaction effort was varied to produce a range of air voids. Thi was somewhat difficult with the Exxon Rolling Wheel Compactor, which produced lower than expected sample air voids. Six tests were used to evaluate the engineering properties of the HMA: indirect tensile strength at 25 ?C, resilient modulus at 0 and 25 ?C, Marshall Stability, Hveem Stability, repeated load cyclic creep at 40 ?C and compressive strength at 40 ?C. Only the indirect tensile strength, resilient modulu 38 sions were used to determine slope and offset values between air ed s r h field nly 10 of 20 cases (50 percent). The numbers of differences between the lly different at the 5 percent significance level. he field cores and laboratory compacted ples lso note that the Texas Gyratory Compactor is ng wheel compactors were. Based on this study, the Texas Gyratory Compactor as recommended for the production of laboratory specimens (34). Based on the AAMAS study, the research conducted by Button et al. (34) and the ted to use a SHRP CG Marshall Stability tests were conducted on samples compacted with the Marshall Hammer; HMA from only two sites were compacted and tested with the Elf Linear Kneading Compactor (34). Linear regres voids (x variable) and the test result (y variable) for the field cores and samples compacted with the various compactors for each site. Statistical analyses were perform to compare the slope and intercepts for a given test between the field cores and sample compacted with each of the laboratory compactors used. The Texas Gyratory Compacto produced samples equivalent to field cores in 24 of 33 cases (73 percent). The Exxon Rolling Wheel compactor and the Elf Linear Kneading Compactor produced samples with equivalent properties to field cores in 18 of 28 and 9 of 14 cases, respectively (bot 64 percent). The Marshall Hammer produced samples with equivalent properties to cores in o different compactors were not statistica The authors note that the differences between t sam were relatively small. They a more convenient, faster and cheaper for producing samples at a given air void level than the rolli w work completed by the French with the PCG, the SHRP researchers elec gyratory compactor for the production of routine testing samples (29). Further, the researchers selected a protocol similar to the French PCG. As noted previously, the P 39 mpact cant r 2. Vertical pressure = 600 kPa (87 psi), 3. Speed of gyration = 30 rpm. The development of the design compaction level, Ndesign will be discussed later in the report. compacts samples at six gyrations per minute. The SHRP researchers desired to co samples as fast as possible to decrease testing time (4 samples compacted to 200 gyrations takes approximately one half day at 6 gyrations per minute). As noted previously, McRae and McDaniel (26) found the effect of gyration rate to be insignifi up to 10 gyrations per minute. Therefore, the SHRP researchers designed an experiment to assess the effect of gyration rate on the resulting volumetric properties of the compacted sample. A single aggregate source and a single asphalt source were used in the experiment. Samples were compacted at optimum and optimum ? 1.0 percent asphalt content. Samples were compacted at 6, 15 and 30 gyrations per minute. Volumetric properties evaluated included optimum asphalt content, air voids, VMA and VFA. Ai void contents of 4.4, 4.5 and 4.0 percent were reported, respectively, for 6, 15 and 30 gyrations per minute. Statistically, these values were not different. Therefore, the SHRP researcher selected a gyration rate of 30 gyrations per minute to minimize testing time (29). The initial characteristics of the SHRP Gyratory Compactor were selected as follows: 1. Angle of gyration = 1 degree, 40 2.2.2 Studies to Evaluate Factors Affecting Gyratory Compaction Prior to the conclusion of the SHRP research program, initial studies were conducted to compare specifications for gyratory compactors and their effect on the resulting sample properties. A study was conducted to compare a SHRP Gyratory compactor, built by the Rainhart Company, a modified Texas Gyratory Compactor and a Corps of Engineers GTM (29). The SHRP Gyratory Compactor could be used to compact both 4-inch and 6-inch diameter samples. The angle on the Texas Gyratory Compactor was adjusted to 1 degree, and a frequency controller was added to allow the compaction speed to be set to 30 rpm. A single aggregate source, binder source, and ere compacted at t content and optimum ? 1.0 percent. Two replicates were compacted in ere nsity at in Figure 2.9. Changes in sample gradation (19.0 mm NMAS) were used for the study. Samples w optimum asphal the SHRP and Texas Gyratory compactors and three replicates were compacted in the Corps of Engineers GTM. A larger study is described to compare the SHRP Gyratory and modified Texas Gyratory, but the results are not presented. Based on the French concept of reporting the log of gyrations (x-axis) versus sample density (y-axis) reported by Moultier (35) in reference (29), three parameter w identified to compare the compactors: C 10 , C 230 and K, where, C 10 is the sample de 10 gyrations, C 230 is the sample density at 230 gyrations, and K is the slope of the densification line. The parameters are illustrated asphalt content are expected to affect the compaction curve as illustrated in Figure 2.10. Figure 2.9. Typical Gyratory Compaction Curve (29). Figure 2.10. Effect of Asphalt Content on Compaction (29). own The results of the experiment to compare the three gyratory compactors are sh in Table 2.1. For the optimum minus samples, the corps of Engineers GTM produced significantly higher sample densities than the SHRP Gyratory at C 10 and all other samples 41 42 at (29) Gyratory Compactor at C 230 . At optimum plus, the compacted sample densities were significantly different C 10 for all three compactors; at C 230 the Corps of Engineers GTM results and 6-inch diameter SHRP Gyratory results were significantly different from each other and significantly different from the other samples. Thus, it was concluded that the different gyratory compactors did not compact the same. TABLE 2.1 Comparison of Densification Parameters from Gyratory Compactors SHRP AC% Parameter 4-inch 6-inch Texas Modified Corps GTM C 10 83.4 84.4 85.4 86.8 C 230 92.0 91.3 92.4 93.7 Optimum Minus K 6.281 5.039 5.100 5.059 C 10 85.6 86.4 87.1 89.0 C 230 95.2 94.4 95.0 96.5 Optimum K 7.100 5.958 5.858 5.531 C 10 88.5 88.8 90.0 91.6 C 230 99.0 98.0 99.0 99.4 Optimum Plus K 7.732 6.772 6.598 5.724 It was observed that the modified Texas Gyratory Compactor had an angle of gyration of 0.97 degrees (external) while the SHRP Gyratory Compactor had angles of 1.14 and 1.30 degrees, respectively, when compacting the 6-inch and 4-inch diameter samples. Cominski et al. (29) concluded, ?A variation in the angle of compaction of ? 0.02 degrees resulted in an air voids variation of ? 0.22 percnt at 100 gyrations.? This difference resulted in a change in optimum asphalt content of ? 0.15 percent. Based on this research, the specification for angle of gyration was changed to 1.0 ? 0.02 degre The differences in compaction with the Corps of Engineers GTM were attributed to the manner in which the angle is indu ss. ced. The angle of gyration for the Corps of Engineers GTM is fixed at only two points, one of which (the oil roller) allows the angle 43 the these two compactors with the modified Texas gyratory ompac search e del d as f gyration, mold loading procedure, compaction d. A to vary if the pressure in the roller is exceeded, while the SHRP and modified Texas Gyratory Compactors fix the angle at three points. In 1994, two models of SGC?s were initially approved as meeting the specifications for the SHRP (now called Superpave) Gyratory Compactor or SGC by FHWA in a pooled fund purchase for state departments of transportation: the Pine Instruments Company (Pine) model number AFGC125X and the Troxler Electronic Laboratories, Inc. (Troxler) model number 4140 (36, 37). A study conducted by the Asphalt Institute (38) compared c tor used to develop the Superpave criteria during the Strategic Highway Re Program and a prototype Rainhart (SHRP) Compactor. Three samples of each of six blends were compacted in each compactor at optimum asphalt content. At Ndesign, th Pine compactor produced similar results to the Modified Texas compactor and the Troxler compactor produced results similar to the Rainhart Compactor. The Pine Mo AFGC125X produced significantly higher densities than the Troxler Model 4140 did in five of six comparisons. After the completion of this study, modifications were made to both the Pine and Troxler SGCs. Subsequently, both the Pine Model AFGC125X and Troxler 4140 SGCs were included in a ruggedness study to evaluate AASHTO TP4 (39). The ruggedness study was conducted according to ASTM C1067. As specified, seven factors were evaluate part of the ruggedness study: angle o pressure, precompaction, compaction temperature, specimen height, and aging perio high and low level was selected for each of these factors. Due to the difficulty in obtaining exact external angles of gyration and exact specimen heights, some tolerance 44 s in he .0 kPa, , ce om ? n t used was allowed for both of these parameters. The low range for external angle of gyration varied from 1.22 to 1.24 degrees and the high angle varied from 1.26 to 1.28 degrees. The specification for the angle of gyration had been changed to 1.25 ? 0.02 degree 1994 during the original Ndesign experiment (29). This will be discussed later in t document. Fixed batch masses of 4500 and 5000 g were used to produce sample heights of approximately 110 and 120 mm. Four 19.0 mm NMAS mixes representing two aggregate types (crushed limestone and crushed river gravel) and two gradations (coarse and fine) were used in the experiment. The range for compaction pressure, then specified as ? 3 percent or ? 18 caused significant differences in three of five laboratories for one or more mixes (4 cases total). Marginally significant differences were found in seven of twenty cases for the height extremes. Additional analysis of the data indicated that the actual differences (approximately 12 mm) exceeded the 10 mm target difference. The 12 mm differen caused marginally significant differences for the fine graded mixes. Therefore, it was recommended that the existing tolerance on sample height in AASHTO TP4 be relaxed fr 1 mm to ? 5 mm (39). The two ranges for external angle of gyration only resulted in a significant difference in one in twenty cases. As anticipated, higher angles did produce denser specimens, but regression analysis indicated that only one percent of the difference i sample density was explained by the change in angle and the relationship was not significant (39). Both compactor types responded similarly to all seven of the main effects. However, additional analyses indicated differences in sample density between the laboratories that used the Pine AFGC125X compactor and the laboratories tha 45 ple r ompactors grouped together (39). ead across the United States, several additio ng fore, 5 mpar the Troxler 4140 compactor. Paired comparisons using a t-distribution grouped the three labs using the Pine compactor together and the two labs using the Troxler compactors together for three of the four mixes with the Pine compactors producing higher sam densities. There were three groupings for the fourth mix, but once again the Troxle c As the use of the SGC became widespr nal manufacturers have developed SGC?s. In addition, both Pine and Troxler have developed new models of SGC?s. This led to the need to develop a means of evaluati the new SGC?s to ensure that they would produce results similar to the Pine AFGC125X and Troxler 4140. AASHTO TP4 did not contain a precision statement (36). There it was not clear what the acceptable difference between various SGCs should be. To address potential differences between compactors, FHWA developed a standard protocol to compare compactors, which was approved by the FHWA Superpave Mixtures Expert Task Group, and is designated AASHTO PP35, ?Standard Practice for Evaluation of Superpave Gyratory Compactors (SGCs)? (36, 37, 40). AASHTO PP3 consists of a comparison between a single unit of the new compactor versus one of the two original pooled fund compactors (Pine AFGC125X or Troxler 4140). The co ison consists of compacting six replicate samples for each of four mixes in both compactors. The mixes specified include: a 12.5 mm nominal maximum aggregate size (NMAS) mix, two 19.0 mm NMAS mixes (one coarse and one fine graded) and a 25.0 mm NMAS mix. The comparison is to be performed at one of the five Superpave Regional Centers (36). When evaluating new models, both Pine and Troxler performed the AASHTO PP35 comparisons against their respective original compactor (37, 41). 46 s in en , coarse-graded mixes and one fine-graded mix were d nal Angle of Gyration ion n the during l to the opposite Many agencies, throughout the country, have reported significant difference the bulk specific gravity of compacted samples from different SGCs, which have be properly calibrated. Iowa Department of Transportation (42) completed a study to address this very concern. They evaluated four brands of SGCs: Pine AFGC125X Troxler 4140, Test Quip Brovold and Interlaken Model 1. Four 19.0 mm nominal maximum aggregate size mixes, three used in the study. All of the compactors were calibrated according to the manufacturer?s recommendations prior to testing. The Troxler compactor was found to produce consistently higher densities at Ninitial. This was believed to be related to the manner in which the angle is induced. The Pine SGC consistently produced the highest density an the Interlaken SGC produced the lowest density at Ndesign. The Interlaken SGC produced the largest differences from the average density of all of the compactors. 2.2.3 Inter The sensitivity of the density of SGC compacted samples to the angle of gyrat was identified during the SHRP (29). The internal angle of gyration is defined as the angle of the interior of the mold wall relative to the top and bottom plates or platens. The platens are assumed to be parallel to one another. The gyration angle (internal and external) changes (generally decreases) with all types of compactors during compaction, primarily due to flexing of the SGC frame, but can be significant with some compactors. One source of compliance is believed to be the ram used to apply vertical pressure o samples. One of the platens is generally attached to the ram. When the ram flexes compaction, the platen supported by the ram may not remain paralle 47 on. l o on easure and display the xternal angle of gyration during compaction based on one to three points. The numerous methods of measuring the external angle of gyration result in a lack of uniformity from one SGC to another. uip Inc., developed an independent device to measure the internal angle of gyration. The device is referred to as the Dynamic Angle Validation Kit (DAVK). The DAVK is placed inside the SGC mold with hot mix asphalt sample. A data acquisition system within the DAVK dynamically records the internal angle of gyration during compaction (43). A draft procedure (40) for evaluating the dynamic internal angle of gyration ?Evaluation of the Superpave Gyratory Compactor?s (SGCs) Angle of Gyration Using the FHWA SGC Angle Validation Kit? was developed by FHWA. platen. For these reasons, the gyration angle must be determined during compaction, preferably with a full-height HMA sample, not in the un-loaded (mold empty) conditi The external angle of gyration is measured differently for each brand and many models (within a brand) of gyratory compactors. The Pine Model AFGC125X uses dia gauges and can measure the static (not gyrating) angle in both the loaded (with a full- height HMA sample) and unloaded condition. The Troxler 4140 uses a digital gauge t dynamically (while the compactor is gyrating) measure the offset of the turntable used to apply the angle in the loaded condition. No means for measuring the angle of gyrati was supplied for the Rainhart compactors. All of the other compactors, Test Quip (Gilson or Pine AFGB1A), Interlaken, Pine Model AFG1A and Troxler 4141, use internal linear voltage displacement transducers (LVDT) to m e The FHWA, in cooperation with Test Q 48 alibrat . ld VK is designed to measure the internal angle of gyration along with a full le, by date the DAVK and a 115 mm tall (final height) HMA sample. This can The DAVK unit is shown in Figure 2.11 with its accompanying NIST traceable c ion standard. The DAVK consists of a machined body designed to fit inside a SGC mold. Two probes connected to a single LVDT protrude through the body and rest against the mold wall. The base of the unit rests against the top or bottom mold plate During compaction, the base of the DAVK is held tightly against the top or bottom mo plate and acts as a reference plane from which the internal angle of gyration is measured using the LVDTs. The DAVK body contains a data acquisition system and power source. The data acquisition system is programmed and the data downloaded to a notebook computer using software provided by the manufacturer. The DA height (115 mm tall) hot mix asphalt (HMA) sample (43). Figure 2.12 illustrates the possible measurements of angle of gyration. The external angle of gyration is defined as ?. The internal angles of gyration are defined as ? T (top) and ? B (bottom) for the angle measured when the DAVK is placed above the HMA samples or below the HMA samp respectively. The measured internal angle of gyration is different when the DAVK is placed at the top or bottom of the mold (43, 44). Therefore, ? T and ? B , as measured the DAVK, should be averaged to determine an effective internal angle of gyration (? AVG ) (43 - 45). The DAVK unit is approximately 77 mm tall. Certain SGC molds cannot accommo be solved by extrapolation (43). Figure 2.11. DAVK and Calibration Block. Figure 2.12. Definition of Internal and External Angle of Gyration. To determine the internal angle of gyration by extrapolation, a series of HMA masses necessary to produce varying height samples are utilized. Typically, three sample masses are used (to produce three different height samples) for the extrapolation for which two replicates of each sample mass are compacted with the DAVK against the 49 50 rch (44, s an excellent linear relationship between sample height and internal angle of gyration with the DAVK at both the top and t tom of the mold. Extrapolations to e then averaged to e of roduced by inducing end plate deflections with machined tapers in the ine AFG1A. Dalton (47) reported on a second study where four compactors, adjusted to the e in xes representing a wide range of NMAS according to the criteria established for AASHTO PP35. Two of the fou K; l- . (48) determined a target DIA of 1.16 degrees. The target was based on setting single articles of the original pooled-fund purchase SGCs, the Pine AFGC125X upper platen and two replicates with the DAVK against the lower platens. Resea 46) indicate he bot 115 mm are performed separately to determine ? T and ? B . ? T and ? B ar produce ? AVG . Studies have been conducted to relate the dynamic internal angle of gyration (DIA) to sample density. Dalton (44) conducted a study to evaluate the effect of DIA on compacted sample using two compactors, the Pine AFGC125X and the Pine AFG1A. Testing indicated that a change in internal angle of 0.1 degrees resulted in a chang 0.014 Gmb units or approximately 0.6 percent air voids for the Pine AFGC125X and a change in internal angle of 0.1 degrees resulted in a change of 0.017 Gmb units or approximately 0.7 percent air voids for the Pine AFG1A. The varying internal angles were artificially p P sam ternal angle of gyration, compared favorably for nine of ten mi r compactors allowed full height HMA samples to be compacted with the DAV one used precompaction and one used extrapolation. The results of this experiment indicated that the measured internal angle of gyration was independent of mix type. FHWA conducted a study to determine the target and tolerance for the DIA. A Khateeb et al 51 and Tro to n er. verage internal angle of gyration versus the average Gmb values by com xler 4140, to an external angle of gyration (using the manufacturer?s calibration equipment) of 1.25 degrees as specified in AASHTO T312, and measuring the DIA using the AVK. Using a 12.5 mm NMAS Superpave mix, the average DIA was determined be 1.176 and 1.140 degrees, respectively for the Pine AFGC125X and Troxler 4140 SGCs. Thus, set at an external angle of 1.25 degrees, the original pooled fund SGCs produced an average DIA of 1.16 degrees. The tolerance was determined to allow a maximum variability of approximately 0.10 percent design asphalt content or 0.25 percent air voids. Using the relationship developed between DIA and Gmb and a target change in air voids of 0.25 percent, the tolerance for DIA was determined to be ?0.03 degrees. Prowell et al. (49) measured the DIA on 112 different SGCs in Alabama (seve different models). Three samples of a 19.0 mm NMAS mix were then compacted to 100 gyrations on each compactor for density determination. Regression analysis using all the data indicated an R 2 = 0.37. This indicates that although DIA explains part of the variability, other factors affect compacted sample density from one laboratory to anoth Figure 2.13 shows the a pactor type for the 19.0 mm NMAS mix at 4.4 percent AC. A simple linear regression was performed with internal angle of gyration as a predictor for Gmb excluding y = 0.0999x + 2.314 R 2 = 0.99 2.44 2.45 2.46 2.48 Rainhart 2.41 2.42 2.43 2.47 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 DIA, degrees Average Gmb Interlaken Troxler 4141 Troxler 4140 Brovold Pine AFGC125X Pine AFG1A On average, a 0.02 degree change in DIA results in a 0.002 in Gmb or 0.08% Air Voids Figure 2.13. G mb versus Average Internal Angle of Gyration (49). the Interlaken and Rainhart data. The R 2 = 0.99 indicates on average an excellent relationship between average internal angle of gyration and average sample bulk density. The relationship shown in Figure 2.13 indicates that on average a change in 0.1 degrees of internal angle will result in a change of 0.010 G mb units or a difference in air voids of approximately 0.4 percent. Therefore, a change of ?0.02 degrees as allowed by AASHTO T312 could produce a difference in air voids of approximately 0.08 percent or based on Superpave?s rule of thumb (all things being equal, a 0.4% change in AC% results in a 1.0% change in air voids) approximately a 0.03 percent difference in design asphalt content. 52 53 first study to relate laboratory compaction to densification under traffic was the Co noted previously, acceler d passes of a dual wheel configuration loaded to 60,000 constructed density was approximated by 98 percent of the density of 50-blow Marshall sample d not on the basis ous - of sphalts, appears to have stabiliz ear. 2.3 DENSIFICATION OF PAVEMENTS UNDER TRAFFIC A number of studies have been conducted to evaluate pavement densification under traffic. Though the general consensus is that pavements reach their ultimate density after the second or third summer, the results in research studies have varied. Additionally, some of these studies have tried to relate in-place density to laboratory compaction. The rps of Engineers Study to develop the Marshall Method (14). As ate loading was used to apply 3,500 passes of a 15,000 lb wheel load; 1,500 37,000 lb wheel load; or 1,500 passes of a lbs to test sections produced at various asphalt contents. It was noted that as- s. The 50-blow compaction effort appears to have been selecte of air voids after traffic, but by comparing the optimum asphalt content obtained with the various compaction efforts to visual assessments of the field performance of the vari sections at different asphalt contents (50). Dillard (51) tracked six Virginia sand asphalt pavements over a 100-week (2 year) period starting in 1952. Coring was conducted 5 times after construction on each the 6 projects. The densification of 4 of the 6 projects, all sand a ed after one year, while the coarser mixes continued to densify in the second y In 4 of 6 cases, 50-blow Marshall samples had a higher density than the pavement did after 2-years of traffic. 54 pled out of ities The authors estimated that between 15 and 20 blows would est ma Twenty additional pavements, 13 HMA and 7 sand asphalt, were sampled the following year. Lift thicknesses ranged from ? to 1 ? inches. HMA was sam haul trucks at the HMA plant and compacted using 30, 50, and 75 blows; sand asphalt samples were compacted with 20, 35, and 50 blows. Cores were taken from each section between 1 to 4 months and between 13 to 16 months after construction. Comparisons were made between the core densities after 13 to 16 months and the Marshall sample densities compacted with the aforementioned blow counts. For the sand asphalt mixes, 30-blows appeared to provide the best correlation with in-place density; for 7 out of 13 sand mixes the mean 30-blow Marshall densities and in-place dens after 13 to 16 months of traffic were not significantly different. Figure 2.14 shows the data for the HMA mixes. b tch the in-place density of the HMA. The authors noted the relative unimportance of traffic in the correlation between number of Marshall blows and in-place pavement density (Figure 2.14). Figure 2.14. 50-Blow Marshall versus In-Place Densities (51) Campen et al. (52) evaluated the densification of pavements placed in Omaha, NE between 1955 and 1959. The pavements were designed with a 50-blow Marshall compaction effort, with maximum aggregate sizes of 1/2, 5/8, and 3/4 inch. Primarily one mix design was used in each year; however, in 1957 the mix was altered from a 5/8 inch to a 3/4 inch maximum size. Laboratory samples were compacted and samples were sawed from the pavement immediately after construction. In 1960, samples were sawed from the pavements at the rate of 4 to 10 per mile. By 1960, 13 of 18 pavements had densified to ? 1.0 percent of the laboratory density, with 3 of those 13 pavements slightly 55 56 does not control ultimate density [this author noted a slight ing the data], ensity obtained from a 50-blow Marshall was not exceeded eling] of the pavement. esistance seemed to have been achieved at the expense of dura slight rutting and shoving at critical exhibited raveling, at times extreme rav rs conclude (52), ?In spite of all the scientific advancement the design as much of an art as it is a science.? This author e to some extent today! acked the densification of 47 test sections on 12 projects through roximately 700 cores and 200 Marsha of adequate traffic data, the authors did not attempt to relate traffic to pavement densification. Instead they presented the average densification of all of the sections with time. They concluded that the pavements densified significantly over the first year, but to a lesser degree over the second year (2.0% average increase in density first year versus 0.6% average increase in the second). Immediately after co nts were less than 95 percent of Marshall density; after on 8 percent and after two years it was reduced to 4 exceeding the laboratory compacted sample density. The authors concluded the following: 1. Ultimate density is achieved in a few months in hot weather, 2. Initial density trend, R 2 = 0.25, when plott 3. The compacted d by heavy traffic, 4. Initial density affects the wear [rav The authors note that rut r bility. The pavements placed in 1955 exhibited locations. Pavements placed after 1955 eling. The autho of bituminous paving mixtures is still believes that statement is still tru Graham et al. (53) tr out New York over a two-year period including app ll samples. Due to a lack nstruction, 29 percent of paveme e year this was reduced to 57 per t of Marshall density would be approximately ict in- m the in ity m ximum size mix with an 85/100 pen binder. The lift thickness was 1 inch. The , and 345 ?F) to produce a range of mix viscosities from approximately 40 to 900 Saybolt Furol Seconds. The sections were cored at the time of construction and 4, 9, and 21 months after construction. Though the as-constructed densities varied, the in-place densities converged under traffic, except for the granite mix placed at 225 ?F and the gravel mix placed at 250 ?F. Binder was recovered from the cores for testing. Initially, the mix cent. [This author notes that 95 percen 91 percent of theoretical maximum density.] An equation was developed to pred place air voids. The three most significant terms were volume of asphalt binder, deflection of the underlying pavement, and deviation of the aggregate gradation fro maximum density line. Woodward and Vicelja (54) monitored the construction of Aviation Boulevard Los Angeles, CA. Two lifts were placed, 3 inches (uncompacted) of 1 ? inch maximum aggregate size base mix and 2 inches (uncompacted) of a ? inch maximum aggregate size surface mix for a compacted thickness of 4 inches. The pavement was cored at the time of construction and 30, 60, and between 90 and 180 days after construction for a total of 169 cores. The average as-constructed density was 133 to 135 lbs/ft 3 . Dens increased approximate 3 lbs/ft 3 in the first 30 days; 1 to 1 ? lbs/ft 3 in the next 30 days; and 1 to 1 ? lbs/ft 3 in the final increment. Permeability tests and a large quantity of other data were collected but not reported. Bright et al. (55) constructed 24 test sections on U. S. Route 64 west of Raleigh, NC. Two coarse aggregates, granite and gravel, were used to produce a ? inch a mixing temperatures in the test sections were altered (225, 250, 287 58 placed at lower temperatures exhibited less binder aging. However, the authors note that by 21 months less binder aging was noted in the sections with higher initial density. Serafin et al. (57) tracked the pavement densification of 6 test sections representing 6 different binder sources (one grade) each subdivided into 5 sub-sections with varying binder content and compaction temperatures on one project in Michigan for 12 years. The pavement was subjected to approximately 8 million tractor-trailer passes during this period. An examination of the reported data indicates the pavement densification leveled off after 4 years of traffic. on of the study conducted by Graham et al. (53) s. The rete yr. ed with Palmer et al. (58) reported on a continuati in New York. The pavement densities were tracked for a period of 5 year authors conclude, ?If such a thing exists as ?ultimate field density? of an asphalt conc mixture, service time to attain this equilibrium may exceed 5 yr. [year] for New York State conditions, whereas studies elsewhere indicate leveling off of density after 1 to 4 [years] of service (ultimate density being defined as that not exceeded with passage of further traffic and/or time).? Epps et al. (58) conducted a study to try and determine the factors which affect the ultimate density of pavements with relation to the laboratory density determin the Texas Gyratory Compactor. The study monitored pavement density on 15 projects in Texas over a two-year period. Based on previous studies, some of which have been discussed in this document, the following factors were suggested as affecting the ultimate pavement density (58): 1) ?Degree of initial compaction 2) Material properties 59 f the pavement is dependent on the ompac a) Aggregate absorption b) Aggregate surface characteristics c) Aggregate gradation d) Asphalt temperature-viscosity relationship e) Asphalt susceptibility to hardening 3) Mix design a) Asphalt content (film thickness) b) Voids in mineral aggregate 4) Weather conditions a) Air temperature variations (daily and seasonal) b) Date of construction 5) Traffic a) Amount b) Type c) Distribution throughout year d) Distribution throughout day e) Distribution in lanes 6) Pavement thickness.? The authors state (58), ?The initial density o c tibility of the mix or the ease with which it can be compacted, the type of compaction equipment, the rolling sequence and procedure, and the timing of the compaction process.? 60 sity e inter will not densify until the onset of warm weather. Little densification as observed during colder months. The authors recommend the use of ESALs to . Figure 2.16 shows ensification as a function of ESALs. The authors concluded that ?Eighty percent of the al effects, was complete within 1 t Cores were taken from the sites after 1 day, 1 week, 1 month, 4 months, 1 year, and two years. Figure 2.15 indicates that pavements compacted to a higher initial den densified less under traffic than pavements compacted to a lower initial density did. Th authors note the importance of season of construction as a pavement constructed in the fall or early w w account for the percentage and weight of trucks in the traffic stream d total 2-year compaction, due to traffic and environment year of service on all of the projects studied.? They also noted that the ultimate pavemen density (for a given project) tended to converge, even if the initial density varied. Figure 2.15. Densification as a Function of Initial Density (58). Figure 2.16. Densification versus ESALs (58). ity and binder properties of 6 pav iod. The densification of the projects app s ontinued on fun easurements and indicate good results when this method was fit to the experimental data. 18 different pavements in 6 states. Thirteen of the cted in the bora ry. T mme unit res by the rm h a t s h d. plott this v e ve s traff an est can be ma of the ount of traffic required to reach the laboratory recompacted density. Kandhal and Wenger (59) tracked the dens ements in Pennsylvania over a 10-year per ear to have leveled off after a 4-year period. However, some densification c three of the projects up until 10 years. The authors suggest the use of a hyperbolic ction to predict ultimate density based on early density m Brown and Cross (60) sampled projects rutted prematurely and 5 performed satisfactorily. The age of the rutted pavements ranged from 1 to 6 years, while the age of the satisfactory pavements ranged from 5 to 16 years old. Cores were tak tes and samples recompaen from the si la to he authors reco nd dividing the in-place weight from co recompacted unit weight to dete ine t e relative moun of den ification t at has occurre By ing alu rsu ic, imate de am 61 62 ed w rs. 5 ta loadings. For low volume roads (average daily traffic of tes Weak trends were noted between the 20 th percentile of the in-place density and the accumulated traffic for both the surface and second layer of the pavement structure. Trends were also observed between the ratio of the in-place unit weight to the laboratory recompacted unit weight versus traffic for both the Corps of Engineers GTM and 75- blow Marshall samples. The best trend (R 2 = 0.50) was for the second lift recompact ith a 75-blow Marshall. Hanson et al. (61) revisited 5-pavement sections that were included in the Asphalt-Aggregate Mixture Analysis System study (32), 5 years after construction. Pavement densities were monitored for a two-year period as part of the original study. A statistical comparison was performed between the measured densities at 2 and 5 yea The comparisons indicated significant differences in 20 out of 30 cases analyzed. As expected, in 16 of 20 cases where significant differences occurred, the air voids after years of service were less than that after 2 years of service. It should be noted that of the 5 projects, 1 was a surface course, 2 were intermediate courses and 2 were base courses. Stroup-Gardiner et al. (62) reported on a 5-year study of 16 projects in Minneso representing a wide range of traffic less than 10,000), the majority of any densification occurred in the first year after construction. For high volume roads, the authors found a decrease in density with time, which they attributed to moisture damage. Brown and Mallick (63) reported on a 3-year study, which evaluated the densification of 6 projects in 5 states. Cores were taken from the projects at the time construction and 1, 2 and 3 years after construction. An examination of the data indica one project reached its ultimate density after 3 years, one project on a very low traffic 63 continued densification has been observed up to 4 and in some cases even 10 years after construction. 2.4 STUDIES RELATED TO Ndesign 2.4.1 Development of the Original Ndesign Table The original Ndesign experiment was conducted by the Asphalt Institute as Task F of SHRP contract A001 (64). The experimental design was primarily developed by the Mixture Design and Analysis System (MiDAS) group consisting of: Ronald Cominski, Gerald Huber, Harold Von Quintus, and Matthew Witczak. The goal of the experiment was to determine the number of gyrations to 1) match the ultimate in-place density, targeted as 96 percent density (Ndesign), and 2) match the as-constructed density, targeted as 92 percent density (Nconstruction). The specifications for the SHRP Gyratory Compactor were discussed previously (29). Sections from the Long-Term Pavement Performance (LTPP) Studies General Paving Sections (GPS) were selected to determine Ndesign and Nconstruction. The in-place density at the time of construction was unknown for the GPS sections, so 92 percent density was assumed. This assumption was not expected to significantly affect the Nconstruction gyrations since only approximately 30 gyrations would be required to obtain 92 percent density. Three hypotheses were identified for the experiment (64): road showed little change and the remaining 4 projects indicated additional increases in density between years 2 and 3. In summary, the literature seems to indicate that the majority of pavement densification under traffic occurs in the first 2 years. However, 64 r and paction (construction and traffic), n between an adjustable compaction parameter of the SGC and the density of the field cores. The experiment was conducted as follows (64): elect s Collect cores and existing data on cores from Material Ref Library 4. mpact, 7. Temperature 1. There was a correlation between lab compaction and field compaction, 2. There was a correlation between lab compaction with the gyratory compacto field com 3. There was a linear correlatio 1. S ites, 2. erence , 3. Separate Cores into paving lifts, Measure bulk specific gravity of each lift, 5. Extract binder and recover aggregate, 6. Remix recovered aggregate with AC-20, short term age, and reco Measure bulk specific gravity and maximum specific gravity of reconstituted mix, 8. Plot densification curves (gyrations versus density), 9. Tabulate and analyze data, 10. Recommend Ndesign values. The experimental matrix is shown in Table 2.2. Two replicates (different pavements) were desired for each cell. The selected pavements were to be at least 12 TABLE 2.2 Experimental Matrix for Original Ndesign Experiment (64) Lift Hot (? 100?F) Warm (? 90 < 100?F) Cool (< 90?F) Traffic Low Medium High Low Medium High Low Medium High Upper X X X X X X X X X L X X X ower X X X X X X 65 n ; medium traffic was defined as greater than 1 million to ss tha n 15 years old to ensure that they had reached their ultimate density. Only single replicates (sites) could be identified for the hot climate. Low traffic was defined as 20-year desig traffic less than 1 million ESALs le n or equal to 15 million ESALs; and high traffic was defined as greater tha million ESALs. The 20-year design traffic was calculated according to Equation 3. ? ? ? ? ? ? ? ServiceinYearsTotal ESALsTrafficdAccumulate , The maximum design traffic included in the experiment was 32.1 million ESALs. Fifteen, 12-inch diameter cores were collected for testing, one from each proje Two 4-inch diameter samples were compacted from e ? ?=ficDesignTrafrYea 2020 (3) ct. ach of the two selected lifts from STM D each project. After completing the first round of compaction, the Asphalt Institute realized that the Rainhart SHRP Gyratory Compactor had erroneously been set to an angle of 1.3 degrees and not the 1 degree angle specified. Therefore, the compacted samples were re-extracted, remixed with virgin AC-20 and recompacted in the Rainhart Gyratory Compactor, now set to an (external) angle of 1 degree. No discussion was provided on the possible effects from aggregate breakdown which may have occurred during the first compaction cycle. It was observed that the sample bulk specific gravities determined with A 2726 were approximately 2 percent higher than those estimated using the SGC sample height and mold diameter [Reference (64) actually says the reverse, but this is an error]. Two gyration levels were picked off of the plots of corrected sample density versus number of gyrations: Nconstruction = 92 percent density and Ndesign = the in-place pavement density. This author notes that the in-place density for two of the lifts, one 66 traffic. erefore, ees. upper and one lower, were less than 92 percent density after more that 12 years of No relationship was observed between traffic and gyrations for the lower lift. Th the determination of Ndesign for the lower lifts was not reported. Figure 2.17 shows a comparison between the Ndesign levels determined at an angle of 1 and 1.3 degr Figure 2.17. Comparison of Ndesign from Angles of 1 and 1.3 Degrees (64) ents, 3 hot, 6 warm, and 6 cool. Linear regressions were erformed between the logarithm (Log) of gyrations and the Log of 20-year ESALs. egressions were performed on the whole data set, and the data set subdivided by limate. One sample, with an in-place density of 99.6 percent, was removed from the 6 c tained after 230 The complete data set consisted of 30 data points representing two gyratory samples from each of 15 pavem p R c warm limate data as an outlier. This level of density was not ob gyrations. The models, subdivided by climate were recommended and are shown below with their pertinent statistical parameters (Table 2.3). The lack of fit statistic was not 67 ates. 2 significant for this model. The climatic zones were redefined as average 7-day high temperatures of 44, 39, and 34 ?C, respectively, for the hot, warm, and cool clim Seven traffic ranges were identified, ranging from less than 0.3 to greater than 100 million ESALs. TABLE 2.3 Ndesign Models (64) Climate Model R ANOVA P-value Hot Ndesign = 10 0.66 0.05 1.34276+0.10850?Log (Traffic, ESALs) Warm Ndesign = 10 0.69 1.26454+0.11206?Log (Traffic, ESALs) 0.00 Cool Ndesign = 10 1.21211+0.09148?Log (Traffic, ESALs) 0.72 0.00 Note: analysis of variance (ANOVA) It is clear that this was a limited experiment. It is noted that the MiDAS group desired to provide the best estimate possible, considering the time available and realized that future research would likely be needed to verify the estimates (64). The next step in the development of the original Superpave Ndesign table was the determination of the numbers of gyrations for Ninitial (then termed N 89 ) and Nmax (the termed N n . 98 ) for each of the traffic levels and climatic zones (29). This was accomplished by translating the original compaction curves horizontally until the density at Ndesign corresponded to 96 percent (Figure 2.17). This translation is based on some of the principles investigated by Moultier (35). The ratio of Log (Nmax) to Log (Ndesign) and the ratio of Log (Ninitial) to Log (Ndesign) was determined for each compaction curve The average ratios were 0.47 and 1.22 for Ninitial and Nmax, respectively. Based on this work, SHRP recommended the following equations (29): NdesignLogNinitialLog ?= 45.0 (4) 68 t was conducted to evaluate the SGC for field control (29). he 2.36 t res and 1993. Cominski et al. (29) state, ?Although the ort for roject. hus the NdesignLogNLog ?= 15.1max (5) The density at Ninitial was specified as less than 89 percent to prevent tenderness during compaction and the density at Nmax was specified as less than 98 percent to prevent rutting at the end of service life. An experimen Changes in asphalt content, percent passing the 0.075 mm sieve, percent passing t mm sieve, NMAS, and the ratio of natural to crushed fine aggregate were experimental variables. A partial factorial experiment was performed. Asphalt content, percen passing the 0.075 mm sieve, and the ratio of natural to crushed fine aggregate all had significant effects on the compaction curve. Based on this experiment, the SHRP researchers recommended the SGC for field control. Finally, the prototype SHRP gyratory compactor was used to design 7 mixtu for nine pilot SPS-9 projects in 4 states: Arizona, Indiana, Maryland and Wisconsin. The sections were constructed in 1992 original gyratory design specified an angle of gyration of 1?, a vertical pressure of 0.6 MPa (87 psi), and 30 rpm, problems were encountered on some SPS-9 mix designs. It became apparent that the 1? angle of gyration provided insufficient compaction eff the air voids required at N design .? An example is provided for the Arizona SPS-9 p The measured density at Ndesign was 90.8 and 92.0 percent, respectively for an (external) angle of 0.97 and 1.27 degrees at trial asphalt content of 4.1 percent. T estimated asphalt content to achieve 4 percent air voids at Ndesign would have been 6.2 and 5.7 percent, respectively, at an (external) angle of 0.97 and 1.27 degrees. It is 69 ation determined by Blankenship (64). The remaining levels appear to be interpolated. TABLE 2.4 Original Ndesign Table (1) Design 7-day Maximum Air Temperature (?C) expected, but not stated, that the specified angle of gyration for the SGC was increased to 1.25 degrees due to concerns about the higher than expected design asphalt contents (29). Table 2.4 presents the original Ndesign table. This author has never seen documentation of the decision to go from the three climatic levels presented by Blankenship (64) to the four levels provided in the original table. The Ndesign gyr levels for the 43 to 45 ?C climate match the gyrations levels for the hot climate < 39 39 - 41 41 - 43 43 ? 45 Traffic (ESALs) N ini N des N max N ini N des N max N ini N des N max N ini N des N max < 3 x 10 5 7 68 104 7 74 114 7 78 121 7 82 127 < 1 x 10 6 7 76 117 7 83 129 7 88 138 8 93 146 < 3 x 10 6 7 86 134 8 95 150 8 100 158 8 105 167 < 1 x 10 7 8 96 152 8 106 169 8 113 181 9 119 192 < 3 x 10 7 8 109 174 9 121 195 9 128 208 9 135 220 < 1 x 10 8 9 126 204 9 139 228 9 146 240 10 3 253 15 > 1 x 10 8 9 143 235 10 158 262 10 165 275 10 172 288 S s were to be compacted to Nmax and the density at Ndesign and Ninitial back calculated using the sample heights recorded by the SGC (Equation 6). ample nGyrationatHeight NatHeight NatDensitynGyrationatDensity max max ?= This is a simplified version of Equations 3-6, 3-7, and 3-8 presented by Cominski (1), produced by combining terms. (6) 70 e completion of SHRP and the release of the Superpave mix design ystem ) and to fe. n conducted on hear tester. The authors applied a factor of 8.97 to e des by 2.4.2 Research Related to Ndesign Conducted after SHRP Following th s , a number of studies have been conducted to compare the results of the Superpave mix design system to previously used design systems (such as Marshall or Hveem refine the Ndesign levels. Sousa et al. (65) report on an early application of the performance based Superpave design on a project on Interstate 17 north of Phoenix, AZ. Two, 1 mile test sections were placed by the Arizona DOT. The mix was a three inch layer of a 19.0 mm NMAS mixture which was to be designed for 10 million ESALs in a 10-year design life. Rutting was to be limited to less than 10 mm over the design li This appears to be the same mix discussed previously by Cominski et al. (29), which resulted in the angle for the SGC being increased from 1 to 1.25 degrees. A fine-graded mixture was selected using a crushed gravel aggregate source with 95 percent one face crushed and 90 percent two face crushed. The mixture was produced with a modified PG 70-10 binder. The optimum binder content was selected based o tests with the repetitive simple shear test at constant height (RSST-CH) test the simple (later called Superpave) s th ign traffic of 10 million ESALs to determine a traffic level of 89.7 million ESALs with 95 percent reliability. Using this traffic level and a plot of asphalt content versus applied ESALs resulting in 10 mm of predicted rutting based on the tests conducted with the RSST-CH, an optimum asphalt content of 4.2 percent was selected. The RSST-CH tests appear to have been conducted at 3 asphalt contents, 4.0, 4.5, and 5.0 percent. By comparison, testing performed with the SGC on field mix resulted in 6.3 percent air voids at an Ndesign of 135 gyrations and 75-blow Marshall compaction effort, then used 71 s (OTA) Mobile Laboratory. The lab conducted tests on four state agency paving projects to demonstrate field control with a prototype SGC. mparisons were performed between SGC and Marshall compacted samples. A unique y the small Arizona DOT, also resulted in 6.3 percent air voids. An optimum asphalt content of 5.2 percent was predicted with the SGC and later verified at 5.1 percent. (This author notes that an optimum asphalt content of 5.0 percent would have been determined using the design traffic of 10 million ESALs (50 percent reliability)). The authors conclude that samples compacted using rolling wheel compactor best match the performance properties of the field cores based on comparisons made with samples compacted in the California Kneading Compactor, Texas Gyratory Compactor, 2 SHRP Rainhart compactors and the Marshall Hammer. Harman et al. (66) reported on testing conducted by the FHWA Office of Technology Application Co relationship was found between SGC and Marshall sample air voids for each project. Ndesign of 100 gyrations produced samples with lower air voids than 6-inch diameter 112-blow Marshall compaction did. The same held true for comparisons between Ndesign of 126 gyrations and 50-blow Marshall and comparison between Ndesign of 113 and 75-blow Marshall samples. Gowda et al. (67) conducted a study to evaluate the sensitivity of volumetric properties and optimum asphalt content to the Superpave Ndesign levels resulting from variations in design traffic and climate. The authors were concerned b differences in Ndesign between some traffic and climate levels (Table 2.4). Four aggregate gradations were selected for the study; all coarse graded (passing below the restricted zone). Two aggregate sources were used in the study: a granite source 72 nd. contents, 4.5, 5.5, and 6.5 percent. 3 with 7 accounted for three of the blends and a sandstone source was used for the fourth ble Two binders were used in the study, a PG 64-22 and a polymer modified PG 76-22. Samples were compacted at three asphalt Three replicate samples of each of the 24 combinations (4 mixes x 2 binders x asphalt contents) were compacted to 288 gyrations (Nmax for > 100 million ESALs a 7-day maximum air temperature of 43 to 45 ?C). The volumetric properties at the 2 Ndesign levels were back calculated from these samples. Figure 2.18 shows the calculated VMA as a function of Ndesign. Note that for a given gradation, VMA changes by approximately 0.3 percent for a change in Ndesign of 10 gyrations. Figure 2.18. Variation in VMA w Statistical analyses were conducted to pare the volumetric properties between 6 gyration levels that only va x design properties for the 4 clim statistically significan cases for optim resulting fro nificant statistical differences were observed in 35 of ith Ndesign for PG 64-22 (67). com ried by 1 to 2 gyrations (e.g. 95 and 96) and the mean mi ates. For the comparison of close gyration levels, t differences were observed 3 of 64 cases for VMA and for 2 of 64 um asphalt content. For the comparison of the different gyration levels m different climates, sig 73 168 cas o VFA. The c which differ design ompared the Superpave and Marshall design procedures for the pr stockpiles. The percen aggregate was held constant and the percen of coarse river sand varied in 5 percent increments duce the 5 blends. All fiv dations were coarse graded. Mixtures ere prepar th an AC-10 (approximately ples were com acted in rshall d to es f r VMA, 2 of 168 cases for optimum asphalt content and 8 of 168 cases for authors concluded that Ndesign levels for differing design traffi by 1 to 2 gyrations do not result in significantly different mix properties and that N levels from differing climates do not result in significantly different mix properties for a given traffic level. Habib et al. (68) c design of shoulder mix in Kansas. Five 19.0 mm NMAS blends were evaluated, oduced from 4 aggregate tage of crushed limestone coarse tage to pro e gra w ed wi PG 58-22). Sam p the SGC to Nmax = 104 gyrations. Volumetric properties were back calculated at Ndesign = 68 gyrations. Four of the five blends, evaluated using the SGC, failed VFA on the low side; the fifth failed dust to effective asphalt content on the low side. Ma samples were compacted with a 50-blow effort for comparison. The Marshall samples met all of the Kansas DOT?s criteria. It was observed that the optimum asphalt contents, VMA and VFA were all lower for the samples compacted in the SGC. The authors speculate that the Superpave Ndesign levels for low volume pavements are approximately 20 percent too high. Mallick et al. (69) reported on the effect on volumetric properties of the restricted zone from mixes produced with crushed and partially crushed fine aggregate and the effect of back calculation on the volumetric properties of samples compacted in the SGC. As discussed previously, when Superpave was first adopted, samples were compacte 74 ratio of the measured Gmb using AASHTO T166 to en Ndesign level, particularly for coarse graded mixes. This e back Nmax and the volumetric properties back calculated at Ndesign. The back calculation uses a correction factor which is the the Gmb calculated with the measured sample mass and estimated sample volume calculated based on the area of the gyratory mold (176.7 cm 2 ) times the sample height recorded by the SGC, cm. Testing conducted with dense and SMA gradations produced with a traprock aggregate indicated that the correction factor varied with the number of gyrations the sample was compacted to. In essence, the sample has more surface texture at lower gyration levels, resulting in a smaller measured volume. Figure 2.19 shows the error in measured air voids. Note that the back calculated air voids are higher than the air voids measured at a giv resulted in a slight reduction in optimum asphalt content for samples compacted to Ndesign as opposed to those compacted to Nmax where volumetric properties wer calculated at Ndesign. Figure 2.19. Error in Back Calculated Air Voids Versus Gyration Level (69). 75 in to 5- SALs and aveme nt ign otes RP Brown and Mallick (63) reported on a preliminary study to evaluate the Ndesign Table. Loose mix, aggregate and asphalt, and cores were sampled from six projects five states in 1992 and 1993. The projects were located in Alabama (2), Idaho, New Mexico, South Carolina, and Wisconsin. The field mix and laboratory mix produced match the field mix were compacted to a number of gyrations which produced approximately 99 percent density with an SGC. Samples were also compacted using 7 blows of a fixed base mechanical Marshall Hammer. A set of 12 cores were obtained at the time of construction and 12, 24 and 36 months after construction. Good correlations were observed between the Log of accumulated E p nt density for 4 of 6 projects. The New Mexico project produced an R 2 = 0.52. This author notes that this may be related to the polymer modified AC 40 used for the project. The remaining projects used AC-20 or softer binders. The one of the two Alabama projects with a poor correlation received very little traffic, approximately 112,000 ESALs after 3 years. On average, the reheated mix was observed to have approximately 1 perce lower density than the laboratory prepared mix did. The difference decreased with increasing gyration levels. The average of the reheated field mix and the laboratory prepared mix were used to estimate Ndesign for each project. The results from one project, I-90 in Idaho, were discarded since it began to rut after two years. The Ndes values from this study predicted to match the in-place density after three years were approximately 30 gyrations less than those determined during SHRP. (This author n that some of this difference might be attributed to the 1 degree angle used during SH 76 : ed implemented, ect (actually one (64), there were t except for the h The authors present a comparison of the Ndesign levels determined in the original Ndesign experiment (Table 2.5) based on Reference (64) for angles of gyration of 1 and otice t t Ndesign is een 27 and 46 gyrations less at an angle of gyration of 1.3 degrees. 70) and the 1.25 degree angle used in this study). The SGC samples had approximately 1.5 percent higher density than the 75-blow Marshall samples. Forstie and Corum (70) performed an initial evaluation of Ndesign for the Arizona DOT. The authors note three concerns about the SHRP Ndesign experiment 1. The angle of gyration used to develop the original Ndesign table was 1 degree, but an angle of gyration of 1.25 degrees was later selected by SHRP without modifying the Ndesign table, 2. The original Ndesign experiment was performed using 100 mm diameter specimens whereas SHRP later specified 150 mm diameter samples, 3. The mixes used in the original Ndesign study were predominately fine grad whereas coarse graded mixes were more predominant when Superpave was first 4. The Ndesign study was based on only two cores per proj wo cores per cell ot climate). 1.3 degrees. N ha betw TABLE 2.5 Comparison of Ndesign Levels for Hot Climate for 1 and 1.3 Degrees ( Predicted Ndesign Design Traffic (Million ESALs) External Angle = 1.30? External Angle = 1.0? 0.5 64 91 3.0 77 111 30.0 97 143 10.0 87 127 77 which had been subjected to 2 to 5 y heel path g the ignition furnace and the aggregate ple re an ced urnace t higher than those later obtained by as of Cores were taken from six in service pavements ears of heavy interstate traffic. The in-place density was determined for the w cores. The asphalt was extracted usin recovered. The actual mix correction factor for the ignition furnace was unknown. The recovered binder was remixed with binder of the same grade as had been used previously and compacted to the appropriate Nmax using a Troxler SGC after which the sam densities were back calculated at Ndesign. The Gmb values for the SGC samples we average of 0.037 units higher or 2.3 lbs/ft 3 higher than the in-place core densities. The SGC densities were also calculated at the Ndesign value for 1.3 degrees. This redu the difference between the laboratory compacted samples to 0.012 Gmb units or 0.7 lbs/ft 3 . Two possible flaws in the study noted by the authors were 1) the ignition f asphalt contents were approximately 0.3 percen solvent extraction, and 2) changes to the recovered aggregate specific gravity were noted resulting from the ignition furnace. Buchanan (71) conducted much of the research which supported NCHRP 9-9, ?Refinement of the Superpave Gyratory Compaction Procedure.? The major objectives of this research were to determine whether, and to what extent, the Ndesign compaction matrix could be consolidated from the original 28 levels determined during SHRP, and secondly to evaluate the back calculation of Ndesign from Nmax. The first objective w evaluated by examining the effect of Ndesign on volumetric properties. An evaluation the parameters of the SGC: gyration angle, vertical pressure, and gyration speed, was not included in this research. 78 r ed; est -22, was d - source, and son for the as made that the average An experimental matrix was developed for the research which included fou aggregate sources, two gradations and six Ndesign levels. The aggregate sources included: New York Gravel, Georgia Granite, Alabama Limestone, and Nevada Gravel. Both gradations were 12.5 mm NMAS; one was fine graded, and one was coarse grad neither passed through the restricted zone. The gyration levels consisted of the low (68) and highest (172) in the original Ndesign table, three intermediate gyrations levels (93, 113, and 139), and 40 gyrations. Based on previous work, it was felt that a lower level of gyrations may be required for low volume roads. A single binder, PG 64 used in the experiment. Three asphalt contents were used to bracket Ndesign. The samples were compacted to Ndesign (not Nmax). Separate samples were compacted to Nmax for three Ndesign levels and compared to results from the Asphalt Pavement Analyzer. Some of the samples did not meet all of the volumetric requirements. The data indicated that optimum asphalt content, VMA, and VFA all decrease with increasing Ndesign; the coarse-graded mixes were more sensitive that the fine graded mixes were. ANOVA was performed to determine which of the experimental factors affected VMA. All of the main factors (e.g., Ndesign, aggregate gradation) and their interactions were significant. Duncan?s multiple range compari procedure was conducted to compare the measured VMA resulting from the differing Ndesign levels. The analyses were conducted separately for the coarse-graded and the fine-graded mixes. For both gradations, the differing Ndesign levels used in this study resulted in significantly different VMA at the 5 percent significance level. An evaluation was performed of the need for the differing gyration levels differing climatic zones in the Ndesign table. The argument w 79 7-d mperature is less than 39 ? United States. Further higher temperatu exist, a stiffer bind ould likely be used. Statistical comp re conducted usi a Student?s t-test between the resulting VM alculated for each aggregate source and gradation between the Ndesign climatic xtrem nd ere evaluated to ign graded ld as te samples could be compacted to Nmax after the optimum asphalt ay maximum te C for the majority of the , where res er w arisons we ng A c e es for a given traffic level (e.g., 68 versus 82 gyrations, respectively for < 39 a 43 to 45 ?C). No significant differences were observed for 41 of 56 comparisons. For the 15 comparisons which were significant, the average absolute difference in VMA was 0.57 percent. Based on these analyses, the differing Ndesign levels as a function of climate were eliminated from the Ndesign table, collapsing the table from 28 to 7 levels. Since the coarse-graded mixes were more sensitive to Ndesign than the fine graded mixes were, the VMA results for the coarse-graded mixes w further consolidate the Ndesign table. The average difference in VMA between Ndes levels was 0.32 percent for the coarse-graded mixes and 0.18 percent for the fine- mixes. A VMA range of 1 percent was selected for differing Ndesign levels. This wou result in a difference in optimum asphalt content of approximately 0.45 percent for the coarse graded mixes. Thus three levels of Ndesign were proposed 70, 100 and, 130 gyrations. A fourth Ndesign level, 50 gyrations, was proposed for low volume roads. None of the mixes included in this study failed the Nmax criteria. Further, it w determined that compacting samples to Nmax and back calculating the volumetric properties at Ndesign can result in errors of up to 0.8 percent air voids. Therefore, it was recommended that samples be compacted to Ndesign for the determination of volumetric properties. Separa 80 content is determined. Table 2.6 presents the revised Ndesign table recommended by Buchanan (71). TABLE 2.6 Revised Ndesign Table Proposed by Buchanan (71) Gyration Levels Design Level ESALs) % Gmm @ Traffic (million Ninitial Ndesign Nmax Ninitial % Gmm at Nmax <0.1 6 50 74 < 91.5 0.1 to < 1.0 7 70 107 < 90.5 > 30.0 9 130 212 < 89.0 < 98.0 1.0 to < 30.0 8 100 158 < 89.0 Anderson et al. (3) conducted an evaluation Ndesign based on the sensitivity of engineering properties to changes in Ndesign. This research had four tasks (originally five, but one was abandoned because it duplicated NCHRP 9-9): 1. Examine the performance of in-place Superpave pavements designed with the original SHRP Ndesign table, 2. Select a validated performance test for rutting, 3. Determine the sensitivity of the performance test to changes in Ndesign, 4. Recommend a new Ndesign table. Six Superpave mix designs were developed using two aggregate types, crushed limestone and crushed gravel, and three Ndesign levels, 70, 100, and 130 gyrations. All of the mixes were 12.5 mm NMAS. The gradations of the three blends for each aggregate source were varied to produce a VMA slightly above the minimum (14.0 rcent). This was done based on the assumption that since binder is the most expensive component of HMA, the mix designers will alter the gradation to reduce VMA as Ndesign decreases. The resulting mixes had measured VMA ranging from 14.2 to 14.6 pe 81 s ted ear 0 ?C. regate. For a given aggregate, there were no significant compactor compared to the other compactors by Consuegra et al. (31) or simply more asphalt in the mixture]. For the RSCH test, the limestone aggregate was again identified percent and optimum asphalt contents of either 4.6 or 4.7 percent. Samples were produced with a single unmodified PG 70-22. The rutting properties of the mixes were evaluated using two tests performed in the Superpave Shear Tester (SST): frequency sweep at constant height (FSCH) and RSCH. FSCH is conducted by applying a small shear stress to the samples which result in a shear strain of less than 0.0005. Tests are conducted at ten frequencies: 10, 5, 2, 1, 0.5, 0.2, 0.1, 0.05, 0.02, and 0.01 Hz. Highway traffic speeds are generally represen by the results at 10 Hz. The complex shear modulus (G*) is the ratio of the applied sh stress to the resulting shear strain. Higher G* values at a given temperature indicate a stiffer mix. FSCH testing was conducted at two temperatures 50 and 60 ?C. RSCH is performed by applying a haversine shear stress of 69 kPa with a 0.1 second load and 0.6 second rest period (1.4 Hz) for 5000 cycles. The test result is reported as the accumulated permanent shear strain after 5000 cycles. Testing was conducted at 6 It was observed that G* (10 Hz) was significantly higher for the limestone aggregate than for the gravel agg differences between the stiffness of the mix designed at 100 and 130 gyrations. G* (10 Hz) was significantly lower for both aggregate mixtures designed with Ndesign = 70 gyrations. There was a general trend of decreasing shear stiffness with decreasing Ndesign. It was believed that this trend is related to changes in the aggregate skeleton. [Alternatively, this author believes it could be related to the degree of contact developed between the aggregate particles, similar to the results observed for the kneading 82 as being more rut resistant. However, no significant differences were noted between the accumulated shear strain from the RSCH test for the mixes designed at different Ndesign levels. A study was also conducted to examine the sensitivity of VMA to Ndesign. Similar results to NCHRP 9-9 were noted. Finally, the authors note that based on experience, an increase in one high temperature binder grade, say from PG 70 to PG 76 will result in the same increase in mix G* as a change of 30 gyrations. In 1999 at a meeting of the FHWA Superpave Mixtures Expert Task Group (ETG), Dr. Ray Brown and Mr. Mike Anderson presented the results of their respective studies on Ndesign. This author was a member of the ETG at that time and present at the meeting. Based on that meeting, a new Ndesign table was recommended and adopted by n below AASHTO in 2001. The revised Ndesign Table from AASHTO PP28 is show (Table 2.7) (36). In 2004, AASHTO PP28 was adopted as AASHTO M323 (4). TABLE 2.7 Superpave Gyratory Compaction Effort (36) Compaction Parameter Design ESALs (millions) Ninitial Ndesign Nmax < 0.3 6 50 75 0.3 to <3 7 75 115 3 to < 30 8 100 160 ? 30 9 125 205 In 1994, Colorado DOT initiated a study to compare the air void contents of laboratory compacted samples and in-place field projects (72). At the time the study was initiated, Colorado DOT was using the Texas Gyratory with variable end-point stresses for the differing traffic and environmental conditions within Colorado. Samples were taken from 25 sites at 22 projects, designed using the Texas Gyratory, and compacted in a 83 e of traffic and environmental conditions. l five to six years. The in-place air e, of traffic. Note from the figure that the mples at 4 percent air voids. Harmelink and Aschenbreber (72) in their mmendations state that the mi b esi s o the env n c n lorado. Tw o e ggest were: 1) lowering Ndesign and 2) adjusting the mix design air void content ess than 4 percent). It is noted that Colorado DOT uses 100 mm diameter molds in the GC, which tend to produce lower density that 150 mm diameter molds would. Pine SGC. The mix designs also met the Superpave design criteria. The projects were selected to cover a rang At the time of construction, loose mix was sampled and 3 samples each were compacted to the specified Ndesign and one level above and one level below the specified Ndesign. Fifteen cores were taken to determine the as-constructed density, 5 from the estimated position of the left-hand wheel path of the design lane and 5 cores just to the right and 5 cores just to the left of the estimated position of the left-hand whee path. All but 3 of the 25 sites fell within the specified in-place density range of 92 to 96 percent, with an average density of 94.7 percent. Five cores were then taken from the left-hand wheel path on an annual basis for a period of void contents from the 3, 4, 5, and, 6 year cores did not change significantly. Therefor it was concluded that the pavements reached their ultimate density after approximately 3 years of traffic. Figure 2.20 shows a comparison between the laboratory compacted air voids at Ndesign and the in-place air voids after 3 years in-place air voids are approximately 1.2 percent higher than the laboratory compacted sa reco xes are eing d gned at too low of an a phalt c ntent for ironme tal and traffic onditio s in Co o opti ns for adjustm nts su ed (l S 8 y = 0.8852x + 1.6638 R 2 = 0.3842 2 3 Voids (3 Ye ) , Field Mixed (FMF C ) 84 0 1 012345678 % Air Voids at Superpave Ndesign with Field Mixed/ Lab Compacted (FMLC) FMFC Voids > FMLC Voids Gyrations too High 4 5 6 7 % Air ars / Field Compac t ed FMFC Voids < FMLC Voids Gyrations too Low Difference in voids, 1.2% Line of Equality d 16 l area. All of the projects were 12.5 mm NMAS. A pavement Figure 2.20. Comparison of Ndesign and In-Place Air Voids after 3 Years (72) Watson et al. (73) conducted a study to verify the Ndesign levels for Georgia Department of Transportation. The objective of this study was to compare the performance of Georgia DOT?s mixes designed using the Superpave and the Marshall mix design systems, both produced using PG binders and aggregates from the same source. From a list of 217 Marshall and Superpave projects, 16 Marshall designed an Superpave designed projects were selected that matched closely in age, traffic, aggregate source, and geographica performance survey and coring was conducted at each site. Three cores were collected from each project, one in each wheel path and one from between the wheel paths. Quality control and quality assurance data were determined from historical records. Figure 2.21 shows a comparison of the in-place air voids in the wheel path. The average 85 ality uperpave projects averaged 6.1 ns of the in-place air voids for the Superpave designed projects were 5.7 percent whereas the in- place air voids for the Marshall designed projects were 3.8 percent. Data from the qu assurance records indicated that the in-place air voids at the time of construction averaged 7.3 and 6.1 percent for the Superpave and the Marshall designed mixes, respectively. It should be noted that the Marshall and S and 4.7 years old, respectively. Figure 2.22 shows a comparison between the design VMA for the Superpave and Marshall designed mixes. The authors note that the average VMA for the Superpave designed mixes (14.9 percent) is almost 2 percent less than the average VMA for the Marshall designed mixes (16.8 percent). This occurred even though the gradatio Marshall designed mixes were closer to the maximum density line than the gradations of Field Air Voids (From Wheelpath) 0.0 1.0 2.0 3.0 4.0 5.0 Project No. 6.0 7.0 8.0 9.0 10.0 % ) 01234567891011213141516 Va ( Marshall Superpave Avg. Superpave = 5.7 Figure 2.21. Comparison of Superpave and Marshall in-place Air Voids (73) VMA Comparison 14 15 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Project No. 16 17 18 VM ( % A ) Marshall Superpave Marshall Avg. = 16.8 fort with the Superpave mix design system. It should be noted that Georgia Figure 2.22. Comparison of Superpave and Marshall Design VMA (73) the Superpave designed mixes were. This indicates the effect of the increased laboratory compaction ef DOT used effective specific gravity to calculate VMA for both the Marshall and the Superpave designed mixes. The difference in design VMA resulted in the average asphalt content for the Superpave designed mixes being 0.34 percent less than that for the Marshall designed mixes. Locking Point Pine (74) proposed the ?Locking Point? concept for the SGC. The locking point was likened to the growth curve conducted to determine the maximum number of roller passes in the field before the increase in in-place density leveled off or decreased. It was noted that mixes are not compacted with the same number of passes in the field because 86 87 t were f nted ht able 2.8. ination (75) Gyration 1 2 3 4 5 6 7 8 9 10 each mix is different. Rolling was stopped at the peak density before excessive aggregate degradation occurred. The locking point concept was developed from comparisons made between three years of Marshall and Superpave data and field growth curves. Initially, the locking poin was defined as the first gyration in a set of three gyrations of the same height which preceded by two gyrations of the same height (0.1mm taller). It was believed to indicate the development of some degree of coarse aggregate interlock and be related to the density achieved in the field growth curves. It was noted that the standard deviation o the number gyrations equal to the locking point was less than the standard deviation of the number of gyrations to obtain 4 percent air voids. Vavrick and Carpenter (75) discuss errors in the back calculated density from samples compacted to Nmax. A refined definition of the locking point is also prese where the locking point is defined as the first gyration in the first occurrence of three gyrations of the same height proceeded by two sets of two gyrations with the same heig (each 0.1 mm taller) as illustrated in T TABLE 2.8 Sample Gyratory Height Data Illustrating Locking Point Determ 60 111.9 111.9 111.8 111.8 111.7 111.7 111.6 111.6 111.5 111.5 70 111.4 111.4 111.3 111.3 111.2 LP 111.2 111.2 111.1 111.1 111.0 80 111.0 110.9 110.9 110.8 110.8 110.8 110.7 110.7 110.7 110.6 88 76. m (5) recognized . ompaction procedure which samples with the same mechanical properties as field-compacted HMA ). The most widely recognized study of this nature was that conducted by the Corps of Engineers during the development of the Marshall mix design procedure. More than 214 test sections representing 27 mixes were placed and tested with accelerated loading. Three wheel loads were used: 15,000, 37,000 and 60,000 lbs; 3500 passes were applied 2.5 SUMMARY OF LITERATURE REVIEW The first HMA (actually sand asphalt) was placed in the United States in 18 Initially, optimum asphalt content was selected by experience. Several proprietary mixes were developed, and widely used. As the popularity of HMA grew, there developed a need for standardized tests to assist with the design and control of HMA. This was partially due to the fact that there were no longer enough experienced individuals to make decisions regarding the adequacy of a mix (5, 6). One of the first tests applied to the determination of optimum asphalt content was the pat test, basically a visual assessment of the residual asphalt on a piece of Manila paper which had been pressed into a fresh sample of HMA (9). Hvee the relationship between aggregate gradation and optimum asphalt content, finer mixes generally require higher optimum asphalt contents because they have more surface area. In the 1930?s researchers began to look for a laboratory compaction procedure which would produce sample densities similar to the ultimate density of the in-place pavement Pavements were observed to densify under traffic for a period of 2 to 3 years or more. Later this search was expanded to include a laboratory c would produce (5, 12, 14, 15, 21, 22 89 with the 15,000 lb load and 1500 passes with the remaining two loads. The filler content and asphalt content of each mixture were each varied at three levels. Based on field performance, optimum asphalt content for each mixture was recommended. The laboratory compaction effort that produced an optimum asphalt content that best matched those determined in the field was 50-blows (14, 15). Hveem (5) placed less emphasis on sample air voids and more emphasis on stability, but did recognize the importance of air voids as they relate to durability. Texas conducted studies with the Texas Gyratory Compactor during the 1940?s to verify that the laboratory compaction effort matched the ultimate pavement density. The density of cores taken 1 to 12 years after construction averaged 0.8 percent lower than the laboratory samples. The Corps of Engineers developed the GTM in response to even higher (up to 350 psi) tire pressures on military aircraft (12, 25, 26). A general summary of the early design philosophies might be that HMA should be designed with the highest asphalt content (for durability) which does not result in stability or rutting problems. Marshall emphasized the importance of minimizing VMA by using the densest aggregate structure possible (6). Numerous studies were conducted to monitor the densification of pavements, in under traffic after 2 to 3 years, with most of the densification occurring in the first year. Some studies observed densification over a longer period of time (up to ten years). Attempts were made to relate field densification to laboratory compaction, particularly with the Marshall method. situ (14, 32, 50 ? 63). Generally, pavements were believed to reach their ultimate density 90 ecame more prevalent in the United States. This is somew d tire pressure on trucks. To address these concerns, 50 million dollars was devoted to asphalt research in the SHRP program authorized by the Surface Transportation and Uniform Relocation Assistance Act of 1987 (1). Superpave was a product of the SHRP research program. The gyratory com ctor w s se cted routine use in the Superpave mix design system for it ability to 1 oduce am ith similar mechanica ope as d com ed HM and 2 r its c nve ence , 31, ). rther e Fr h in ted a relationship between the number of gyrations and the layer thickness and number of onal characteristics of pactor were adopted, with the exception that the speed of gyration was increased to An experiment was conducted during SHRP to determine Ndesign (29, 64). The ment f s a linear clim n in service halt extracte own and ass etween pavement density and traffic for the lower lifts (> 100 mm); therefore these samples were not tested (64). In the late 1970?s and 1980?s, rutting problems b hat attributed to the use of radial tires and increase pa a le for ) pr s ples w l pr rties fiel pact A, ) fo o ni (29 34 Fu , th enc dica roller passes in the field. The operati the French Gyratory Com 30 rpm (28). premise of the experiment was three-fold, 1) there was a relationship between pave densification and accumulated traffic, 2) there was a relationship between the densities o samples compacted in the SGC and in-place density, and 3) there wa relationship between Ndesign and design traffic. Fifteen pavements representing three atic regions and three traffic levels were cored (one core each) which had bee for more than 12 years. The density of the cores was measured and the asp d to recover the aggregate. The density at the time of construction was unkn umed to be 92 percent. No relationship was observed b 91 The 230 gy sity was ba and Ndesig s 1.3 degrees not compac yration of 1 degree. From this a table of Nde the SHRP gyratio ime eve and agencies compared the resu systems they were familiar wi m enerally produce lo re lower optimum asphalt contents than the Marshall sy Res c exist between mix pro Errors w ax, as originally r to Ndesign (69 t research was conducted to confirm these findings which resulted in levels and a change in practice fro c Ndesign to sim mples to Ndesign for volumetric property recovered aggregate was remixed with virgin asphalt and two samples compacted to rations for each mix. The number of gyrations which matched the in-place den ck calculated. A relationship was developed between design traffic (ESALs) n. However, it was found that the angle of gyration of the SGC wa the specified 1.0. Therefore, the aggregates were again recovered, remixed and ted in the SGC, now set to an angle of g sign levels for three climates and 7 traffic levels was developed (29, 64). Later researchers expanded this table to 4 climates (29). Late in SHRP, the angle of n was changed to 1.25 degrees. The Ndesign levels were not altered at this t n though angles had been demonstrated to affect Ndesign (29). When Superpave was first released, researchers lts from the Superpave system using the SGC to the design th, ost frequently the Marshall system. The SGC was found to g wer VMA, air voids and therefo stem did (63, 66, 68, 70). ear h indicated that significant differences did not perties resulting from many of the Ndesign levels which were close together (67, 71). ere observed between the density at Ndesign back calculated from Nm ecommended in the Superpave system, and the density of samples compacted , 71). Significan a consolidation of the Ndesign table from 28 to four m ompacting samples to Nmax and back calculating volumetric properties at ply compacting sa 92 determinat sensitivity of v and performance test results related to rutting of 5 to 6 OT ly d ot on n ion. However, the consolidation of the Ndesign table was primarily based on olumetric properties laboratory produced mixtures, not relationships with field performance (2, 3, 71). Colorado DOT conducted a study that indicated that in-place air voids after years of traffic were higher than those obtained at Ndesign using the SGC. Lower design gyrations or design air void contents were recommended (73). A study for Georgia D indicated that the design VMA of 12.5 mm NMAS Superpave mixes was approximate 2 percent less than Marshall designed mixes with corresponding aggregate sources (74). Illinois DOT developed the locking point concept to prevent the over compaction of an subsequent aggregate degradation in the SGC. The locking point was believed to be related to the maximum achievable density during construction (75). The literature indicates that there is still concern that the Ndesign levels have n been optimized to maximize field performance. The original Ndesign table was based a limited data set for which the as-constructed densities were not available. The Ndesig table was consolidated based on a laboratory study design to evaluate the sensitivity of volumetric properties to Ndesign. There is a need to verify the current Ndesign values and relate them to field densification and performance. 93 CHAPTER 3 RESEARCH TEST PLAN 3.1 RESEARCH TEST PLAN However, this consolidation was based on the sensitivity of both volumetric properties and a performance test related to rutting to Ndesign; it was not tied to field performance. There is still concern that the Ndesign levels, in some cases, may be too high. Two states have adopted a single gyration level to design mixes; one of these has been successfully with respect to field performance. In order to validate the Ndesign levels, an extensive field study was conducted to relate Ndesign to the in-place densification of pavements under various traffic loadings while monitoring field performance. The approach selected for this study was similar to 63 ject included: Ndesign level, lift thickness relative to NMAS, gradation and PG binder grade. The original experimental plan is shown in Table 3.1. Forty projects were required to fill the experimental plan. The projects were geographically distributed across the United States as shown in Figure 3.1. Attempts were made to identify projects in the southwestern and northeastern United States. Projects in the southwest were typically overlaid with open-graded friction course and therefore not suitable for the study. Projects could not be identified in the northeast that could be sampled during the In 1999, the Ndesign table was revised and consolidated from 28 to 4 levels. used for more than four years. Therefore there is a need to validate the Ndesign levels the approach used by Brown and Mallick ( ). Experimental variables for the pro 94 tudy Lift thickness / nominal maximum aggregate size TABLE 3.1 Test Plan for Field Densification S 2 3 4 High T erature Perform nce Grade emp a G atio Level Fine or Coarse Graded Normal m +1 +2 Nor +1 yr n +1 +2 Nor al mal +2 F X X X 50 C X X X F X X X 75 C X X X F X X X X X X X 100 C X X X X X X X F X X X X X X X 125 C X X X X X X X Figure . Locati f F Pr ects 3.1 on o ield oj . 95 required timeframe. In 2000, twenty-two projects were visited and samples were obtained and tested. In 2001, the remaining eighteen projects were visited and samples obtained and tested. All of the mixes sampled were surface mixes. For each project, the following testing and evaluation procedure was conducted: 1. Samples of loose mix were sampled from a truck at the asphalt plant; the corresponding location where the remainder of the mix was placed on the roadway was marked. Where possible, three samples were taken from each project, but in some cases only two could be obtained, 2. Three replicate pills (gyratory samples) were compacted to two different gyration levels, 100 and 160, without reheating, using two different SGCs in a mobile laboratory. This resulted in the compaction of 12 SGC pills per production in sample, or 24 to 36 pills per project, 3. Samples were split and boxed for determination of maximum specific gravity (Gmm), asphalt content and gradation, 4. Three cores were taken from the right wheel path of the area marked on the roadway where the mix corresponding to a given sample was laid. This resulted 6 (2 samples) to 9 (3 samples) cores per project at the time of construction, 5. Gyratory pills, cores, and loose mix for Gmm and asphalt content and gradation testing were brought back to NCAT for testing, 6. The following tests were run at the NCAT laboratory: a. Compacted sample specific gravity (Gmb) by ASHTO T166, b. Gmm by AASHTO T209, c. Asphalt content determination by AASHTO T164, 96 n d. The cores were shipped back to NCAT for specific gravity determination above. project. Brown and Ma pacted SGC sample density of reh to each site so that the SGC samples could be compacted without reheating. Previous res and models of er Model 4141, we e bac ll possibl minimize ed to allow additional coring after four yea hed their ultimate density. The as used at the four-year interval. The d. Washed gradation analysis by AASHTO T30, 7. The sites were revisited at approximately 3 months, 6 months, 1 year, and 2 years after construction. During each visit the following was conducted: a. Three additional cores were taken corresponding to each sample locatio at each project, b. The pavement condition was visually assessed. c. Rut depth measurements were taken adjacent to each core location with a 6-foot string line, as described Mix design and traffic information were also collected for each llick (63) indicated a difference between the com eated and laboratory prepared mix. Therefore, a mobile laboratory was mobilized earch indicated differences in compaction between different brands SGCs (42, 49). Therefore two SGCs, a Pine Model AFG1a and a Troxl re selected for the study. Although previous research had identified errors with th k calculation procedure (69, 71), it was deemed impossible to compact samples to a e Ndesign levels. Two levels, 100 and 160 gyrations were selected to the number of gyrations for which the sample density needed to be back calculated. After two years, the project was extend rs. This was done to ensure that the pavements had reac same procedure as described in No. 7 above w 97 coll e , and SGC compacted sample density information was use n and field performance. ect d traffic, in-place density d to evaluate the relationship between Ndesig 98 AP R ST S S A LY S PR C E TE A summar th oje ele fo e study is shown in Table 4.1. The data le 4 g ed d n le co po g t , 75 00 25 ns. in h c or e data are ted igh temperature binder grade and al sig ve he ibu o tor Ta .1 ides e ting s e u f S rpa t the time p ts sa d. S al wer u he t cur n l . O on oje as tif wit n N n o ons, o wi de of gyr ns 86) pro cts w Nd f yrati 0 ), roj wi d of gyrations. Although only one ith an N n 0 g ion as led, it wi s t t tio e ra meets the intent of the experim tal d gn. Th fferent N S sa ed: (1 oje 12 26 cts) (3 projects). T ver lif kn wa term ed f the average of the easurem at tim f co uc . F - an oars aded xes epa by ir p nt sin e 2 mm ve n ded es a as g de pe t p g .3 sieve above (finer than) the um ity . se ed s are defined as having the design percent g th m iev lo oa tha e imu den lin igur CH TE 4 TE RE ULT AND NA SE 4.1 OJE TS S LEC D y of e pr cts s cted r th in Tab .1 are roup by N esig vel, rres ndin o 50 , 1 and 1 gyratio With eac ateg y, th sor by h bumps actu Nde n le l. T distr tion f fac s in ble 4 prov som interes note on th se o upe ve a the rojec were mple ever states e still sing t original Ndesign levels. These projects were grouped with the closes rent Ndesig evel nly e pr ct w iden ied h a desig f 50 gyrati 12 pr jects th N sign 75 atio (68- , 17 je ith esign o 100 g ons (9 -109 10 p ects th N esign 125 project w desig of 5 yrat s w samp ll be hown later tha he distribu n of d sign t ffic en esi ree di MA were mpl 9.5 1 pr cts), .5 ( proje and 19.0 mm he a age t thic ess s de in rom core thickness m ents the e o nstr tion ine d c e-gr mi were s rated the erce s pas g th .36 sie . Fi e-gra mix re defined havin the sign rcen assin the 2 6 mm maxim dens line Coar -grad mixe passin e 2.36 m s e be w (c rser n) th max m sity e. F e 4.1 99 TABLE 4.1 Summary of Project Information LTPP Grade Grade Used Proje c t ID Roadway NMAS Avg. Thick., mm Lift/ NMAS Fine o r Coarse Graded Neat or Mod. High Low High Low High Temp. Bump Ndesign KY-1 CR 1796 9.5 31.2 3 C N 64 28 64 22 0 50 NE-1 Hwy 8 12.5 39.8 3 F N 64 28 64 22 0 68 KY-3 CR 1779 9.5 27.1 3 F N 64 28 64 22 0 75 MI-2 Hwy 50 9.5 39.9 4 F N 58 28 58 28 0 75 MI-3 Hwy 52 9.5 32.4 3 F N 58 28 58 28 0 75 UT-1 Hwy 150 12.5 38.7 3 F M 64 22 64 34 0 75 NE-3 Hwy 8 12.5 51.2 4 F N 64 28 64 22 0 76 CO-2 Hwy 82 12.5 53.3 4 F M 64 28 64 28 0 86 CO-5 Hwy 82 12.5 44.3 4 F M 64 28 64 28 0 86 AL-5 Hwy 167 12.5 33.7 3 C N 64 16 67 22 0.5 75 FL-1 Davis Hwy 9.5 34.3 4 C N 64 10 67 22 0.5 86 CO-1 Hwy 9 19 49.6 3 F N 52 34 58 28 1 68 CO-4 Hwy 13 12.5 47.6 4 F N 58 34 64 28 1 86 NE-2 Hwy 77 19 48.7 3 F N 64 28 64 22 0 96 MO-2 Hwy 65 12.5 78.8 6 C N 64 22 64 22 0 100 AL-6 Andrews Rd 19 33.0 2 F N 64 16 67 22 0.5 95 AL-2 Hwy 168 12.5 43.1 3 C N 64 22 67 22 0.5 100 100 IL-1 I-57 9.5 40.5 4 C M 64 28 70 22 1 90 0 IN-1 Hwy 136 12.5 44.1 4 C N 58 28 64 22 1 100 100 100 9 100 KY-2 I-64 9.5 33.9 4 C M 64 28 76 22 2 100 100 CO-3 I-70 12.5 50.6 4 C M 64 22 76 28 2 109 AL-4 Hwy 84 12.5 54.1 4 C N 64 16 67 22 0.5 AL-1 Hwy 157 12.5 43.2 3 C N 64 16 67 22 0.5 106 IL-2 I-64 9.5 44.5 5 C M 64 22 70 22 1 9 KS-1 I-70 9.5 22.3 2 F M 64 28 70 28 1 TN-1 Hwy 171 12.5 34.8 3 F M 64 22 70 22 1 IL-3 I-70 9.5 45.7 5 C M 64 28 70 22 1 105 NE-4 I-80 12.5 55.2 4 F M 64 28 70 28 1 10 AL-3 Hwy 80 12.5 38.0 3 C M 64 10 76 22 2 100 GA-1 Hwy 13 12.5 44.1 4 F M 64 16 76 22 2 WI-1 I-94 12.5 36.3 3 C M 58 28 70 28 2 IN-2 I-69 12.5 37.1 3 C N 58 28 64 22 1 125 MI-1 I-75 9.5 35.6 4 C N 58 28 64 22 1 125 64 16 76 22 2 125 64 16 76 22 2 125 AR-3 I-40 12.5 52.8 4 C M 64 16 76 22 2 125 AR-4 I-30 12.5 56.8 5 C M 64 16 76 22 2 125 NC-1 I-85 12.5 45.8 4 F M 64 16 76 22 2 125 MO-1 I-70 12.5 51.1 4 C M 64 22 70 22 1 125 MO-3 I-44 12.5 48.4 4 C M 64 22 70 22 1 125 AR-1 I-40 12.5 53.5 4 C M AR-2 I-55 12.5 51.0 4 C M 0 14 100 2 4 6 2345 Lift Thickness/NMAS Number 8 10 12 6 of Projec ts Fine Graded Coarse Graded Figure 4.1. Frequency Distribution of Lift Thickness to NMAS by Gradation. illu ed mixes. r coarse ot exactly imental design, it does indicate a representative distribution of field practice. The Ndesign of 75 gyration projects were p Two-thirds of the Ndesign of 100 gyration projects were coarse-graded and all but one of the of Therefore from this data set, it appears that hi The cli e for each project was determined using LTPPBind Version 2.1 (7 climatic binder sed on the project. As expected, high temperature binder strates the distribution of lift thickness to NMAS ratio for the fine- and coarse-grad From Figure 4.1 it can be seen that there is a trend for thicker lift thicknesses fo graded mixes. Although the distribution of lift thickness to NMAS ratio does n match the exper redominantly fine-graded. Ndesign 125 gyration projects were coarse-graded. gher gyration mixes are more likely to be coarse graded. matic binder grad 7). The high temperature grade bumps were determined by comparing the grade with that u 101 bum identifi the Ndesig Therefo agencie r than th her station data. Binder bum 20-year 4.2 1. 2. the field, 3. Data fr 4. 5. only occurred when the air temperature exceeded 28 ?C. To address the hypotheses, test results are provided as they relate to the following: ata, 2. Estimation of traffic at various sampling intervals, 3. Evaluation of densification under traffic, ps were predominantly found with higher Ndesign levels. Only two projects were ed with Ndesign of 100 gyrations that did not include a binder bump and all of n of 125 gyration projects included at least one high temperature binder bump. re, for design traffic levels greater than 3 million ESALs, the majority of state s included in this data set are using high temperature binder grades that are stiffe e recommended climatic grade based on the LTPP weat ps are recommended for slow moving traffic (less than 70 km/hr [44 mph]) and for design traffic volumes greater than 30 million ESALs (4). TEST RESULTS There are several important hypotheses for this project: Pavement densification is related to traffic, The laboratory design density should match the ultimate density in Therefore, The laboratory compaction effort should be related to traffic. om the 2000 NCAT Test Track (78) supports other hypotheses: Binder grade, particularly modified binders, effects the rate of densification, Densification (the majority of the ?rutting? which occurred at the 2000 NCAT Test Track) 1. Evaluation of the validity of the d 102 . Verification of Ndesign, 5. Evaluation of the locking point concept. 4.2.1 Comparison of Mixture Data to Design Job Mix Formula Table 4.2 presents the job mix formula (JMF) gradation and asphalt content for each of the 40 projects. No JMF was available for project MI-1, constructed as a warranty project. Three solvent extractions were performed for each sample taken at each project according to AASHTO T164, resulting in 6 to 9 extractions per project ng on whether 2 or 3 samples were taken. Washed gradations were performed on the rec ine extracti two or three samples, respectively, were averaged for compar ons for the percent passing the 2.36 mm sieve. The 2.36 mm sieve is one of the control sieves f o the figure representing ? 4.5 percent from the job m represent typical allowed variability for the average of three sam KY-2, MI-2, NE-2 and UT-1, exceeded the ?4.5 percent tolerance on the 2.36 mm sieve. Figure 4.3 shows the design versus average field gradations for the percent passing the 0.075 mm sieve. Lines have been added to the figure representing ? 1.1 percent of the job mix formula, a typical tolerance for three samples for the percent passing the 0.075 mm sieve. The average percent passing the 0.075 mm sieve for fourteen projects exceeded the 1.1 percent tolerance. Four projects exceeded the tolerance by a large amount, CO-5, MO-2, and UT-1. Generally, dust content is expected to increase during production. However, only six of the fourteen 4 dependi overed aggregate according to AASHTO T30. The results from the six to n ons, representing ison with the JMF. Figure 4.2 shows the design versus average field gradati or Superpave mixes. Lines have been added t ix formula, chosen to ples. Four projects, 103 Percent Passing TABLE 4.2 Design Gradation and Optimum Asphalt Content (JMF) Project 19 12.5 9.5 4.75 2.36 1.18 0.6 0.3 0.15 0.075 Design ID AC% AL-1 100 96 79 45 32 25 19 11 6 3.4 4.9 AL-2 100 99 86 47 30 20 15 9 5 3.4 5.3 AL-3 100 90 75 47 34 22 14 7 4 3.0 5 AL-4 100 93 78 47 34 25 19 12 6 4.3 3.65 AL-5 100 99 87 57 36 25 18 12 7 4.2 5 AL-6 99 87 78 66 49 38 25 14 7 4.6 5.25 AR-1 100 96 78 45 31 21 15 11 7 4.8 5.1 AR-2 100 93 83 40 29 22 16 13 9 5.4 4.9 AR-3 100 94 83 46 30 20 15 12 8 5.6 5.5 AR-4 100 95 84 55 37 25 18 11 7 4.6 5.5 CO-1 99 89 78 59 44 31 22 15 11 7.4 6.1 CO-2 100 96 85 60 45 34 24 17 11 7.6 5.5 CO-3 100 94 81 57 35 24 17 13 9 6.4 5.6 CO-4 100 100 89 56 36 27 20 NA NA 6.5 5.3 CO-5 100 96 85 60 45 34 24 17 11 7.6 5.5 FL-1 100 100 97 65 40 29 23 14 9 5.3 5.7 GA-1 100 98 85 NA 38 NA NA NA NA 5.0 4.8 IL-1 100 100 99 59 32 22 16 9 5 4.3 5.5 IL-2 100 98 90 57 34 22 14 9 7 5.5 5.5 IL-3 100 100 98 57 36 23 14 9 6 4.9 5.33 IN-1 100 100 91 59 39 NA 15 NA NA 6.0 6.4 IN-2 100 100 95 58 43 NA 20 NA NA 3.9 5.6 KS-1 100 100 90 54 38 25 17 11 7 5.0 5.7 KY-1 100 100 95 69 41 27 19 10 NA 5.0 5.8 KY-2 100 100 98 67 39 25 18 11 NA 4.5 5.8 KY-3 100 100 94 69 46 31 21 8 5 4.5 5.6 MI-1 MI-2 100 100 100 83 63 40 28 19 10 5.7 6.8 MI-3 100 100 100 80 55 41 31 19 10 5.0 6.2 MO-1 100 97 85 49 29 17 10 6 4 3.1 5.5 MO-2 100 98 83 48 31 18 13 10 8 6.7 6 MO-3 100 98 89 52 28 18 12 9 7 5.7 6 NC-1 100 95 89 58 43 33 23 14 9 5.4 5.1 NE-1 100 95 90 78 49 30 23 12 NA 3.6 5.5 NE-2 99 90 81 62 41 27 19 11 6 3.4 5 NE-3 100 90 81 71 50 32 25 12 NA 3.5 5.3 NE-4 100 91 87 73 51 34 23 14 NA 6.1 4.8 TN-1 100 98 86 58 43 32 22 10 5 4.0 5.1 UT-1 100 100 89 70 62 45 31 15 NA 6.8 5.4 WI-1 100 98 90 62 39 26 17 9 5 3.5 5.1 104 10 10 20 30 40 50 60 70 80 20 30 D 4 rce 0 nt P 5 g 2 0 .36 60 e 70 80 esign Pe assin m S g. Fi el d Percent Passi ng 2 . 36 mm Sie v e NE-2 m iev A v UT-1 KY-2 MI-2 Lines Represent +/- 4.5 ples d Percent Passing the 2.36 mm Sieve. % Tolerance for Avg. of 3 Sam Figure 4.2. Design versus Average Fiel 2 3 4 5 6 7 8 9 10 1034567 Avg. Fi el d Perc en t Pas s ing 0. 075 mm Si eve MO-2 Design Percent Passing 0.075 mm Sie CO-5 UT-1 ve 89 Figure 4.3. Design versus Average Field Percent Passing the 0.075 mm Sieve. 105 projects exceeding th e high side. Figure 4.4 shows the design versus the average recovere phalt contents for the field samples. Lines were added to the figure representing the job mix for hree samples. With one exception, the fou olvent e tions were performed, which may produce lower asphalt contents (incomplete co gnition furnace any agencies now use. Liquid asphalt ls most expensive component in hot ount allowable by the specifications. Fig x rec is a the least am A v e r a g e F i e l d A C % trac ov m rt e 1.1 percent tolerance exceeded it on th d as ? 0.33 percent asphalt from ula, a typical tolerance for the average of t een projects that fell outside of this range were all on the low side. S ery) mpared to the i that m o the mix asphalt; contractors may tend to put in 3.0 3 4. .5 0 5.0 6.5 3.544.555. 66.577.58 Design AC% 4.5 5.5 6.0 7.0 7.5 8.0 3 e 4.4. D ve s Av ie halt Content. ur esign rsu erage F ld Asp 106 Initially, traffic a io ling intervals was estimat vidin ES rted by en e d rio en ng d tim constru Th d uc egrees of err life ent ding grow e use he tr raffi upda reflect th e h i To st po e traffic e , u e De verage annual da fic ( fo r th n w constructed. De ow In ase wt a by agency. In other wa AADT data at The growth rate was fit using a least squares approach and Mi xc Solver routine. iADAD wh AADT N = ed af s b a n p c d a ar of inter DT C = AADT in the year th e s re rowth r 3. Determine the percent trucks. Some agencies measure a combined percentage of all trucks. Other agencies track separate percentages for single units (such as cube trucks) and multiple units (such as tractor trailers). Percent trucks or heavy 4.2.2 Estimation of Traffic t the var us samp ed by di g the design ALs repo the ag cy by th esign pe d and th multiplyi by the elapse e since ction. is metho can prod e varying d or early in the of the pavem depen on the th rat d for t affic. T c data were ted to e actual traffic lev ls during t e monitor ng period. obtain the be ssibl stimates the following proced re was us d: 1. termine a ily traf AADT) r the yea e sectio as 2. termine a gr th rate. some c s the gro h rate w s provided the cases it s fit from historical using Equ ion 7. crosoft E el?s N ) N C AT = T 1(* +A (7) ere, Predict AADT ter N year , N = num er of ye rs betwee when the roject was onstructe nd the ye est, AA e pavem nt was con tructed (or paved), i = g ate. 107 5. Determine directional distribution and lane distribution factors. Directional less AADT values were for a single direction or the agency recommended a specific value. Agency ed for the lane distribution factor. If none were ) TABLE 4.3 Lane Distribution Factors commercial vehicles were recorded as either a single percentage or as percent single units and multiple units. Multiple units generally represent vehicles with predominantly tandem axles except for the steer axle. 4. Determine a truck factor(s) to convert heavy vehicles to ESALs. In some cases agencies used a standard factor for either all trucks or separate factors for single and multiple units. In other cases agencies recorded the AASHTO vehicle classification or single and tandem axles load spectra. In these cases, a truck factor was calculated by multiplying the percentage of total repetitions in a load group by the corresponding equivalent axle load factor for that load group to determine a composite single unit factor and multiple unit factor. distribution was generally assumed to be 0.5 un recommendations were us provided, the recommendations provided in the AASHTO Design Guide (79 were used. Number of Lanes in each Direction Percent of 18-kip ESALs in Design Lane 1 100 2 80-100 4 50-75 3 60-80 6. The accumulated ESALs at each sampling period, as well as the ESALs for the specified design period are calculated according to Equation 8 or 9. NLDTFTiAADTAADTESAL N ??????+?+= 365%2/))1(( (8) CC 108 ) (( ) NLD MTSTMFMTSFSTiAADTAADTESAL N CC ???? ??+?+??+?+= 365 %100%%2/))1(( (9) where: AADT CF?%) he pavement was constructed (or repaved), SF = single unit truck factor to convert to ESALs, MT% = percent multiple unit trucks, MF = multiple unit truck factor to convert to ESALs, and CF = car factor to convert to ESALs Table 4.4 summarizes the factors used to calculate the traffic at various sampling periods. Using the data in Table 4.4, the design traffic at the design interval specified by the agency and the accumulated traffic at each coring interval were calculated. The accumulated traffic at each coring interval was calculated using the actual dates that the coring occurred and not the targeted intervals, e.g. three months, six months, one year, two years and four years. The accumulated or design traffic for each of these intervals is shown in Table 4.5. C = AADT in the year t i = growth rate, N = number of years (or fraction) between construction and sampling time, T% = percent trucks, TF = truck factor to convert trucks to ESALs D = directional distribution factor, L = lane distribution factor, ST% = percent single unit trucks, Lanes Bo Directions Rate Single Units Combo Units Distribution Factor Factor Factor Unit ESAL Unit ESAL ESA Factor De Pe ( TABL Project ID E 4.4 Roa F dwa act y or s u Nu se mbe d t r of o th Ca AA lcu DT lat G e ro Ac wth cu mu % Trucks la ted E % SALs a % t Var D io irec us tion Int al erval L s ane Distribution Co ES mbi AL ned Single Factor Combo Factor Car L sign riod Yrs) AL 7 % 0.99 20 -1 Hwy 157 4 450 2.5% 20.0 0.5 0.95 AL-2 Hwy AL-3 Hwy AL-4 Hwy AL-5 Hwy -2 CO-1 Hwy -2 -3 CO-4 Hwy CO-5 Hwy -1 -1 -1 -2 -2 -3 168 2 7077 2.5% 10.7% 1 0.99 80 4 10870 2.5% 19.0% 0.5 0.9 0.99 84 2 7120 2.8% 14.0% 0.5 1 0.99 167 2 3796 2.5% 10.0% 0.5 1 0.99 AL-6 Andrews Rd 2 1066 3.5% 2.5% 0.5 1 0.99 AR-1 I-40 4 31000 2.4% 27.6% 14% 5.3% 0.5 0.9 1.163 3.77 2 AR I 2 AR-3 I-40 AR-4 I-30 4 22750 5.1% 47.8% 27.4% 10.2% 0.5 0.9 1.163 3.77 0.00 9 4 22193 1.9% 4.3% 0.5% 0.5 0.9 0.249 1.087 CO H CO I 13 2 2279 1.8% 15.3% 10.8% 0.5 1 0.249 1.087 0.00 82 4 15893 2.0% 4.4% 2.0% 0.5 0.9 0.249 1.087 FL Davis Hwy 5 37100 3.0% 2.0% 0.5 0.24 0.89 GA B IL-1 I-57 IL-2 I-64 4 23100 3.0% 4.8% 31.6% 0.5 0.9 0.36 1.32 20 IL-3 I-70 4 19900 3.0% 9.1% 34.2% 0.5 0.9 0.36 1.32 20 IN U 20 IN I 20 KS-1 I-70 4 5461 3.5% 27.8% 1 0.88 0.69 20 KY-1 CR1796 2 211 5.2% 7.9% 0.5 1 0.47 20 KY I- 20 KY C 20 0.5 20 20 20 20 20 20 20 20 20 10 10 10 10 10 20 20 20 0.000 0.000 0.000 -55 4 4 32000 33000 4.7% 5.9% 33.7% 51.8% 19.3% 29.7% 7.2% 0 11.0% 0.5 .5 0.9 1.163 1.163 3.77 0.9 3.77 2 02 3 3 0.003 3 3 0.0004 0.0004 0.0004 0.00 0.00wy -70 B 82 us. 4 6 158 125 93 81 2.0 1.5 % % 4 2 .4% .6% 2 0 .0% .8% 0.5 1 0.9 0.6 0.24 0.24 9 9 1 1 .08 .08 7 7 0.00 uford H wy 4 4 13924 17700 1.6% 3.0% 8.3% 2.3% 1 0.9 0.9 0.97 0.36 1.32 23.7% 0.5 S 1 -69 36 2 4 140 302 80 50 2.5 2.3 % % 2.1 27.3 % % 0 0 .5 .5 1 0.9 1.30 1.30 64 R17 79 4 2 145 26 00 2 2.1 4.8 % % 18.7 7.7 % % 1 1 0.467 0.5 1.07 0.64 109 B 4 a s alculate Accumula Intervals (Continued) oject D Nu of th AA G h ck % in n % Com Unit Directional Distribution Factor Lane Distribution Factor Combined ESAL Factor Single Unit ESAL Factor Combo Unit ESAL Factor Car ESAL Factor Design Period (Yrs) bo s TA LE .4 F ctor used to C Pr I Roadway mber Lanes Bo Directions ted ESALs at Various DT rowt Rate % Tru s S U gle its MI- I 5 8 61 -7 05 % 1 0.8 0.72 20 00 2.2% 5.0 MI-2 Hwy 2 MI-3 Hwy 2 MO-1 I-70 4 1 MO-2 Hwy 65 4 1 MO-3 I-44 4 3 NC-1 I-85 4 6 NE-1 Hwy 8 2 NE-2 Hwy 2 NE-3 Hwy 8 2 NE-4 I-80 4 TN-1 Hwy 2 UT-1 Hwy 2 WI- I 4 6 8 50 % 0.5 1 0.61 20 52 % 0.5 1 0.59 20 85 % 0.5 0.95 1.00 20 94 % 0.5 0.95 1.00 20 27 % 0.5 0.95 1.00 20 13 1.0% 24.0% 0.5 0.8 0.30 1.15 0.00 20 7.7% 12.3% 0.5 1 0.25 0.89 0.00 20 77 4.9% 13.1% 0.5 1 0.23 0.91 0.00 20 4.2% 6.8% 0.5 1 0.25 0.89 0.00 20 6.2% 52.8% 0.5 0.9 0.14 1.01 0.00 20 171 8800 7.7% 2.3% 0.5 1 0.44 1.08 0.002 20 150 1013 14.9% 9.5% 4.1% 1 1 0.55 0.36 0.56 0.0201 20 1 -9 1428 6.8% 1 0.4 0.72 20 5500 7900 00 00 50 46 700 2623 1320 7506 1.5% 1.5% 1.9% 5.3% 2.5% 2.6% 1.5% 1.4% 0.7% 3.6% 4.87% 3.0% 2.0% 8.7 7.6 34.9 9.8 35.8 1 110 111 Project Roadway 3 6 12 24 48 20 Year TABLE 4.5 Accumulated ESALs at Sampling Intervals ID Months Months Months Months Months Design ESALs AL-1 Hwy 157 69,600 129,022 263,972 559,853 1,149,977 6,748,142 AL-2 Hwy 168 34,215 69,022 138,140 296,338 611,855 3,610,001 AL-3 Hwy 80 97,881 170,357 346,635 767,236 8,861,352 AL-4 Hwy 84 58,977 101,426 182,573 402,633 4,899,406 AL-5 Hwy 167 18,854 34,981 65,784 149,147 1,809, AL-6 Andrews Rd. 1,939 2,960 5,323 9,916 19,907 143 AR-2 I-55 942,469 1,562,429 2,957,818 6,590,986 11,850,476 AR-4 I-30 578,939 1,201,114 2,596,098 6,261,493 11,603,641 97,890,077 CO-2 Hwy 82 27,654 76,585 91,905 185,961 385,731 1,01 CO-3 I-70 Bus. 14,675 26,863 48,805 98,324 202,528 52 CO-4 Hwy 13 19,805 36,273 65,592 132,764 274,968 CO-5 Hwy 82 26,897 75,056 90,370 184,395 384,096 1, FL-1 Davis Hwy 8,117 16,784 30,420 62,813 811,658 GA-1 Buford Hwy 133,892 287,006 435,998 798,627 1,568,426 8,803,521 IL-1 I-57 252,510 449,723 948,145 1,963,241 3,970,500 26,28 IL-2 I-64 445,196 792,900 1,671,661 3,461,359 7,000,327 46,34 IN-1 US 136 28,199 41,039 73,589 144,256 372,269 1,850,992 KS-1 I-70 85,315 227,911 374,505 729,765 1,435,783 10,075 KY-2 I-64 181,101 278,340 539,117 1,016,831 2,061,494 12 KY-3 CR1779 857 1,334 2,608 4,988 10,412 MI-1 I-75 211,625 419,507 650,039 1,426,667 2,893,187 15,96 MI-2 Hwy 50 24,456 32,399 54,261 119,143 240,447 1,25 MI-3 Hwy 52 26,258 45,341 0 132,171 278,594 1,515,2 MO-1 I-70 493,003 884,139 1,306,076 2,541,928 4,778,697 27,546,0 MO-2 Hwy 65 107,389 224,065 349,533 734,786 1,462,700 12,517,6 MO-3 I-44 597,842 1,307,458 2,063,169 4,337,141 8,453,012 53,683,94 NE-1 Hwy 8 4,441 10,481 16,872 37,057 67,176 383,385 NE-3 Hwy 8 4,183 10,424 17,010 37,683 68,179 365,719 TN-1 Hwy 171 25,738 58,918 98,776 207,136 428,119 3,490,39 UT-1 Hwy 150 8,014 14,873 27,347 55,992 122,456 771,982 675 ,958 AR-1 I-40 690,394 1,131,450 2,110,407 4,619,146 8,120,222 48,726,562 91,370,805 AR-3 I-40 956,294 1,936,956 4,141,677 9,974,122 18,576,489 170,842,507 CO-1 Hwy 9 20,866 38,064 68,695 138,927 287,854 756,789 7,593 3,624 720,911 017,593 5,917 4,297 IL-3 I-70 365,925 699,160 1,541,346 3,256,535 6,648,086 44,466,336 IN-2 I-69 688,995 957,471 1,827,656 3,586,718 9,265,105 45,150,555 ,962 KY-1 CR1796 530 819 1,591 3,038 6,357 53,706 ,438,605 84,028 6,398 0,146 00 07 75 1 NC-1 I-85 692,210 1,427,287 2,889,164 6,040,907 12,565,156 73,918,507 NE-2 Hwy 77 16,728 39,363 63,672 140,411 255,199 1,450,960 NE-4 I-80 166,950 413,599 671,010 1,529,367 2,841,721 20,084,248 3 I-1 ,748 W I-94 345,088 494,711 597,614 1,316,468 2,557,478 14,614 112 of tionship between traffic and pavement densification at such low traffic levels. he ar Figure 4.5 shows a distribution of the 20-year design traffic for the projects sampled. Although the original experimental matrix was not evenly filled due to the availability projects, Figure 4.5 indicates a good distribution of 20-year design traffic. There are only three projects with less than 300,000 ESALs, however, it is expected that there is not a strong rela There are 21 projects with design traffic between 3 and 30 million ESALs. Under t current AASHTO M 323, all projects with a design traffic level between 3 and 30 million ESALs would be designed with an Ndesign of 100 gyrations (4). The maximum 20-ye design traffic in the SHRP Ndesign experiment was 32.1 million ESALs (1). Nine projects in this study had 20-year design traffic in excess of 30 million ESALs. 0 1 2 3 6 300,000 1,000,000 3,000,000 10,000,000 30,000,000 > 30,000,000 20-Year Design ESALs N u b er of Pro j e Figure 4.5. Distribution of 20-Year Design Traffic. 4 5 7 8 0 m c t s 1 9 113 he in-place density of HMA may be the single factor that most affects the performance of a properly designed mixture (30, 80). A mediocre mix, well constructed with good in-place air voids, will often perform better than a good mix that has been poorly constructed (30). In-place density, between 92 and 97 percent of Gmm for surface mixes passing through or above the Superpave defined restricted zone will generally provide good performance (80). To limit permeability concerns, in-place density greater than 93 to 95 percent of Gmm may be required for larger nominal maximum aggregate size mixtures, stone mastic asphalt or coarse graded Superpave mixtures (81). In-place air voids that are too high may result in permeability to water and excessive binder oxidization, resulting in moisture damage, cracking or raveling (80, 82, 83). In-place density in excess of 97 percent of Gmm may result in permanent deformation or loss of skid resistance (84). Table 4.6 summarizes the average in-place densities for the projects e - 4.2.3 Pavement Densification T at each of the sampling intervals through 2-years; the complete data are presented in th Appendix Table A.41 through A.80. The average in-place as-constructed density for the 40 projects was 91.6 percent. Figure 4.6 shows a cumulative frequency distribution of the average in-place density for the 40 projects at the time of construction. From Figure 4.6, it is evident that 55 percent of the projects had in-place densities less than 92 percent of Gmm and 78 percent of the projects had in-place densities less than 93 percent of Gmm. This indicates that the in place densities of the majority of the projects were less than desired. There may be a number of reasons for the as-constructed in-place densities being less than desired, 114 mm TABLE 4.6 Average In-Place Densities for Field Projects Average In-Place Density, Percent GProject Roadway ID Construction 3 month 6 month 1 Year 2 Year AL-1 Hwy 157 88.7 93.2 93.6 93.0 93.9 AL-2 Hwy 168 88.3 90.3 90.2 90.2 91.8 AL-4 Hwy 84 88.4 92.8 93.1 92.6 94 AL-5 Hwy 167 89.7 93.6 93.8 93.1 94.6 AL-6 Andrews Rd 91.8 93.1 92.7 93.1 93.3 AR-1 I-40 92.0 93.1 93.5 94.1 94.2 AR-2 I-55 89.4 90.9 91.4 91.8 91.8 AR-3 I-40 91.5 94.6 94.8 94.8 94.7 AR-4 I-30 90.9 94.2 93.5 CO-1 Hwy 9 93.8 96.9 AL-3 Hwy 80 89.7 92.8 93.2 93.3 93.6 .3 94.5 94.5 96.5 97.2 98.1 CO-2 Hwy 82 94.7 96.6 96.6 96.9 97.1 6.0 95.6 95.7 CO-4 Hwy 13 93.7 93.3 92.8 94.2 94.2 3.7 94.2 93.8 FL-1 Davis Hwy 91.8 94.2 94.8 94.3 95.2 KS-1 I-70 89.9 91.2 92.1 93.6 93.6 KY-2 I-64 92.2 93.2 93.3 93.9 94.1 CR1779 92.6 93.1 93.7 94.3 94.2 MI-1 I-75 91.3 92.1 92.8 93.4 94.8 96.8 96.8 H 93.7 NA MO-1 I-70 9 MO-2 Hwy 65 9 MO-3 I-44 93.5 94.4 95.3 90 91 93 93.4 92 95 95 95.7 3.0 95.2 95. .3 95.7 1.0 94.8 95. .0 95.4 2.2 94.9 95. .7 97.2 91 93 94 94.3 91 93 N 93.7 5 92 93 94 94.3 CO-3 I-70 93.5 94.6 9 CO-5 Hwy 82 91.6 93.6 9 GA-1 Hwy 13 95.0 95.7 95.8 96.0 96.5 IL-1 I-57 91.0 93.9 93.8 94.2 94.4 IL-2 I-64 91.8 94.2 94.1 94.4 95.2 IL-3 I-70 92.2 94.3 93.9 94.4 94.5 IN-1 Hwy 136 91.3 90.3 90.3 62.3 93.5 IN-2 I-69 91.4 90.7 91.7 94.7 94.1 KY-1 CR1796 85.5 87.3 86.7 87.7 88.5 KY-3 MI-2 Hwy 50 93.1 95.2 96.1 MI-3 wy 52 93.0 94.5 1 96.5 93.4 92.6 96.4 94.2 95.6 92.7 94.3 5.8 4.4 96.5 95.1 95.6 NC-1 I-85 .1 92.8 .7 .0 NE-1 Hwy 8 .6 95.4 .5 .3 NE-2 Hwy 77 9 0 95 NE-3 Hwy 8 9 1 95 NE-4 I-80 9 2 96 TN-1 Hwy 171 150 .1 93.1 .1 .1 UT-1 Hwy .9 93.5 .2 A 2 WI-1 US 4 .4 93.8 .8 .4 1 1-Year not taken n id with pla x se t, N esearc gineer c cores 2 Sectio overla nt-mi al coa CAT R h En elected not to take 1-Year ores. 0 10 20 30 40 50 60 70 80 90 10 85 86 89 90 91 9 94 95 55 0 84 87 88 92 3 96 Gmm, % C u mula tive F r e que ncy, % % 78% 4 ulative F enc trib f As-C ructed e De y. in S y specifications, The compactabilit e m 3. e S uts Figure .6. Cum requ y Dis ution o onst , In-plac nsit includ g: 1. tate agenc 2. y of th ix, The compaction effort or method of compaction used by the contractor, or 4. A combination of these factors. An ANOVA was conducted using the General Linear Model (GLM) to examin factors which may have affected the as-constructed density. The two to three samples from each project were used as replicates, each sample represented by average of three cores. Agency, gradation (coarse or fine), high temperature PG, lift thickness to NMA ratio, and 2000 Ndesign level were considered as factors. 2000 Ndesign level is the Ndesign rounded to the levels adopted in 2000 (50, 75, 100, and 125). The factor inp 115 116 to project, KY-1, was designed at 50 gyrations. The average as- nt As n lthough many agencies have switched (or switched back) to density specifications based on cores sine the implementation of Superpave, Colorado DOT uses the nuclear auge to determine in-place density. Gauges are calibrated to cores at the beginning of the project and density is monitored with additional cores throughout the project. Both the contractor and the agency conduct nuclear density tests. Georgia DOT will adjust the asphalt content of a mixture in the field to ensure in-place density requirements are met. The main effects for lift thickness to NMAS ratio indicated some unexpected trends when agency was included as a factor. It was believed that this may have been due to interactions which could not be analyzed with the replicates available. Therefore, the are summarized in Table 4.1, presented previously. There were insufficient replicates evaluate interactions, particularly considering the 16 levels for agency. Two factors were significant at the 95 percent confidence level: agency and Ndesign. The fitted means for the main effects indicated very low in-place density resulting from mixes with Ndesign of 50 gyrations. Only one constructed density for KY-1 was 85.5 percent. There were no in-place density requirements in the specifications for KY-1. Therefore, this project was eliminated from the date set. The ANOVA was re-run resulting in agency being the only significa factor (p = 0.000). Examination of the main effects indicated that three agencies achieved particularly good as-constructed densities: Colorado, Missouri and Georgia. noted previously, Colorado DOT uses 100 mm diameter SGC molds, which tends to result in lower sample densities and therefore higher asphalt contents which may aid i field compaction (73). All of the Colorado DOT projects used crushed gravel for the coarse aggregate, which may be easier to compact than crushed stone aggregate. A g ANOVA was rerun as described previously without using agency as a factor. Only high temperature PG was significant (p = 0.000); however, this is driven by the PG 67-22 which was only used by two agencies, one of which consistently had low as-constructed densities. The fitted model is poor (R 2 = 0.37) without agency as a factor (R 2 = 0.67 with agency). 117 The main effects plot for the fitted means is shown in Figure 4.7. With the exception of the PG 67, the trends are as expected: increasing density with increasing lift thickness to NMAS, decreasing density with increasing Ndesign level, and increasing density with fine-graded as compared to coarse-graded mixes. As noted previously, coarse-graded mixes tend to require higher in-place density to be impermeable to water (81). Main Effects Plot (fitted means) for Construction Density 65432 93 92 91 7670676458 M e a n o f C o i o ns i t y , % ns t r u c t n D e 90 12510075 93 92 91 90 FC Lift Thickness to NMAS High PG Grade 2000 Ndesign Gradation Figure 4.7. Main Effect Plot for Factors Affecting As-Constructed Density. 118 tween d 6 , on occurred during the winter months (78). The in- een 1 and Figure 4.8 shows a cumulative frequency plot for in-place density for the sampling periods through 2-years. Individual plots, for each project, are shown in the Appendix. From Figure 4.8, it is apparent that the majority of the densification occurs in the first 3 months after construction (63 percent). There is little if any difference be the 3 and 6 month in-place densities. This is most likely due to the fact that projects constructed during the summer would be experiencing cooler weather between 3 an months after construction. This matches the findings from the 2000 NCAT Test Track which indicated that little densificati place density representing the 50 percent frequency increased slightly from 93.0 to 93.2 percent between 6 months and 1 year, and then 1.4 percent to 94.6 percent betw 2 years. 0 10 20 30 80 9 99 In-Place D nsity, unulat e Fre 40 50 60 70 90 100 110 C iv quenc y , % Construction 3 Month 6 Month 1 Year 85 87 8 91 93 95 97 e % Gmm 2 Year mulative Frequency lot fo lace y by l riodFigure 4.8. Cu P r In-P Densit Samp ing Pe 119 slight increase in betw 1 an e lin ls it was impossible to know he ts ha ed their u timate density fter 2-years. The literature suggests that pavements reached their ultimate density after 2 to 3 years of traffic (21, 51, 52, 58), but could densify for a longer period of time (57, 59). Since the goal was to determine the Ndesign gyrations that produced samples with the same density as the ultimate density on the roadway, it was decided to extend the monitoring of the in-place density and take an additional set of cores after 4-years of traffic. The pavement condition survey conducted at the 4-year interval would also provide a better indication of the long-term performance of the pavement. Table 4.7 compares the 2-year and 4-year pavement densities for each project. The average in-place density for all of the projects after both 2- and 4-years was 94.6 percent. Two tests were conducted to compare the 2-year and 4-year pavement densities, Student?s t-test and a paired Student?s t-test. In addition, an F-test was determine whether the model with equal or unequal sample variances should be used. The t-test was used to compare the population means: H 0 : average 2-year density = average 4-year density, H 1 : average 2-year density ? average 4-year density. Whereas the paired test examined the difference between the 2-year and 4-year density at each core site. In three cases, KY-1, NE-2, and NE-3 the F-test indicated that the sample variances were different between the 2-year and 4-year densities. The Student?s t-test for unequal sample variances was used for these sites. The two-tail p-value is reported in all cases. Since there was a density een the d 2 y ar samp g interva if t pavemen d reach l a conducted to compare the sample variances prior to running the Student?s t-test to 120 % Gmm Paired t-test Population t-test TABLE 4.7 Comparison of 2-Year and 4-Year Densities Proje Roadway ? = 0.05? t ? = 0.05? ct 2-Year 4-Year p-value Significant p-value Significan AL-1 Hwy 157 93.9 94.3 0.0886 No 0.2977 No AL-2 Hwy 168 91.8 91.7 0.8968 No 0.9219 No AL-3 Hwy 80 93.6 AL-4 Hwy 84 94.3 AL-5 Hwy 167 94.6 AL-6 Andrews Rd 93.3 93.6 0.1202 No 0.4757 AR-1 I-40 94.2 94.2 0.2629 No 0.6918 AR-2 I-55 91.8 92.1 0.0941 No 0.4186 AR-3 I-40 94.7 94.6 0.7531 No 0.8442 No No No No CO-4 Hwy 13 94.2 94.4 0.4504 No 0.4613 No FL-1 Davis Hwy 95.2 IL-1 I-57 94.4 94.6 0.2052 No 0.5548 No IL-3 I-70 94.5 94.6 0.2154 No 0.5249 No IN-2 I-69 94.1 94.8 0.0735 No 0.2087 No KY-1 CR1796 88.5 87.7 0.5281 No 0.4321 No KY-3 CR1779 94.2 94.4 0.4772 No 0.7774 No MI-2 Hwy 50 96.8 97.4 0.0091 Yes 0.3408 No Yes 0.1508 No MO-1 I-70 96.5 NA MO-2 Hwy 65 95.1 95.0 0.8276 No 0.8836 No C-1 I-85 93.4 93.9 0.0062 Yes 0.0660 No NE-1 Hwy 8 95.7 95.5 0.3002 No 0.6646 No NE-2 NE-3 Hwy 8 95.4 95.2 0.6303 No 0.6330 No TN-1 Hwy 171 94.3 93.6 0.0056 Yes 0.0427 Yes o WI-1 US 45 94.3 94.2 0.6521 No 0.8412 No AR-4 I-30 94.5 94.7 0.0894 No 0.3701 No CO-1 Hwy 9 98.1 97.7 0.1063 No 0.3565 No CO-2 Hwy 82 97.1 96.8 0.0196 Yes 0.4763 No CO-3 I-70 95.7 95.7 0.6190 No 0.8492 No CO-5 Hwy 82 93.8 93.3 0.0645 No 0.3068 No GA-1 Hwy 13 96.5 96.3 0.3201 No 0.6385 No IL-2 I-64 95.2 95.3 0.0265 Yes 0.4559 No IN-1 Hwy 136 93.5 94.1 0.3286 No 0.3541 No KS-1 I-70 93.6 93.0 0.1085 No 0.2985 No KY-2 I-64 94.1 94.4 0.0277 Yes 0.4279 No MI-1 I-75 94.8 94.4 0.0944 No 0.1827 No MI-3 Hwy 52 96.5 96.8 0.0279 1 MO-3 I-44 95.6 95.5 0.6249 No 0.7958 No N Hwy 77 95.7 95.9 0.1870 No 0.3923 No NE-4 I-80 97.2 97.4 0.0268 Yes 0.1964 No UT-1 Hwy 150 93.7 93.6 0.7387 No 0.7850 N 1 Incorrect layer tested on four-year cores (Novachip added between 2- and 4-years). 121 2- iability. The analyses indicate e inal 2-year ed that the ultimate density was achieved It ates a after which time The 4-year density was less than the 2-year density in 15 of 35 cases. If the year and 4-year densities are not different, e.g. the 2-year density is the ?ultimate? density, then lower values would be expected due to testing var that the paired t-tests were significantly different (? = 0.05) in 8 cases, and th average 4-year density was higher in 6 of those 8 cases. However, the paired t-test could be subject to differences due to variances in the longitudinal density of the pavement; although, generally pavement density is believed to be less variable in the longitud direction than in the transverse direction over short distances. The t-test to compare population means was only significantly different (? = 0.05) in one case, TN-1. The average 4-year in-place density (93.6 percent) for TN-1 was less than the average density (94.3 percent). One possible explanation for this could be the onset of moisture damage. Based on these analyses, it is conclud after 2-years of traffic. Factors affecting pavement densification are of interest in this study. Figure 4.9 through Figure 4.11 show typical examples of the observed pavement densification with time. A figure for each project is shown in the Appendix. Figure 4.9 shows the densification of project CO-4. Project CO-4 is a relatively low volume pavement with 20-year design traffic less than 1 million ESALs and a posted speed limit of 55 mph. Figure 4.9 indicates that project CO-4 shows little densification with time or traffic. should be noted that CO-4 was compacted to a relatively high as-constructed density (93.7 percent). Figure 4.10 shows the densification of project AL-1. AL-1 indic significant increase in density in the first 3 months after construction, CO-4 92.0 93.0 94.0 95.0 98.0 avem t D n s i % m m 96.0 97.0 0 50 0 00 10 0 00 0 15 000 0 2 0 000 0 25 000 0 30 000 0 ESALs P en e t y, G ESAL Constructed 8/11/2000 s Figure 4.9. Densification of Project CO-4 with Time and Traffic. the rate of densification levels off. The 20-year design traffic for AL-1 is 6.7 million ESALs. Project AL-1 was compacted to a low as-constructed density. AL-1 rapidly densified to an acceptable level in the first three months. Relatively little densification is observed after the first three months. This may be due to an increased rate of binder oxidization due to the low initial density. Figure 4.11 shows the densification of project MI-1. Project MI-1 is a high volume interstate with a 20-year design traffic of 16.0 rate of densification million ESALs. The higher traffic volume appears to cause a steady up until the 2-year sampling interval. The as-constructed density of project MI-1 was close to typical specifications. These examples demonstrate some of the apparent effects initial density and traffic can have on densification. These will be investigated in greater detail later in the report. 122 AL-1 88.0 89.0 90.0 91.0 92.0 95.0 0 500000 1000000 1500000 ESALs P a v m e nt D e i t y , G 93.0 94.0 e ns % m m ESAL Constructed 4/18/2000 s Figure 4.10. Densification of Project AL-1 with Time and Traffic. MI-1 123 91.0 91.5 92.0 ESALs 92.5 94.0 0 1000000 2000000 3000000 4000000 P avem e n t s % 93.0 93.5 94.5 95.0 D e n i t y, G m m ESALs Constructed 7/17/2000 Figure . 4.11. Densification of Project MI-1 with Time and Traffic 124 months nvestigated. The 3-month ensification was calculated as the difference between the 3-month and as-constructed in- ors which may . 000 l he results with Since the largest percent of pavement densification occurred in the first three , the factors affecting the 3-month densification were i d place density. An ANOVA was conducted using the GLM to examine fact have affected the densification after 3 months. The two to three samples from each project were used as replicates, each sample represented by average of three cores Gradation, high temperature PG or bump in high PG, lift thickness to NMAS ratio, 2 Ndesign level, and month of construction were considered as factors. 2000 Ndesign leve is the Ndesign rounded to the levels adopted in 2000 (50, 75, 100, and 125). High temperature PG bump was considered as an alternate to High PG to better account for climatic differences between the sites. Month of construction was added based on speculation that pavements constructed in the fall would densify less than pavements constructed in the summer would. The factor inputs are summarized in Table 4.1, presented previously. T of the analysis using high temperature PG bump are shown in Table 4.8. High temperature PG bump (p = 0.016) and month of construction (p = 0.000) were identified as significant factors at ? = 0.05. A plot of the main effects is shown in Figure 4.12. The trends are generally as expected. There is a slight trend for increasing densification increasing lift thickness to NMAS, except for the 6:1 ratio. Recall that there is only one project, MO-2, constructed at the 6:1 ratio. Densification decreases with high PG bump (1 grade bump would correspond to a 6 ?C increase in high temperature PG), except for the half-grade bump resulting from the use of PG 67-22. As discussed previously, PG 67-22 was used by only two agencies, one of which tended to have low 125 of Sum of Mean statistic value ?=0.05 TABLE 4.8 ANOVA (GLM) Results for 3-Month Densification Source Degrees Freedom Adjusted Squares Adjusted Squares F- p- Sign.? Lift Thickness to NMAS 4 4.910 1.227 0.86 0.490 No High Temperature PG 3 16.491 5.497 3.86 0.012 2000 Ndesign 3 1.257 0.419 0.29 0.830 No Gradation 1 0.437 0.437 0.31 0.581 Total 109 Bump Yes Month of Construction 6 59.405 9.901 6.95 0.000 Yes No Error 92 131.141 1.425 M e a n o f 3- M nt h D e ns i c a t i o % G m m o f i n, 65432 2.01.00.50.0 1251007550 3 2 1 0 10987654 3 2 1 0 FC Lift Thickness to NMAS High Temp Bump 2000 Ndesign Month Const Gradation Main Effects Plot (fitted means) for 3-Month Densification as-constructed density. Projects with low as-constructed density would be expected to densify more under traffic. Ndesign is neutral except for 50 gyrations. As noted previously, only one 50 gyration project was sampled, KY-1, with no in-place density specifications and a very low as-constructed density. This suggests that the current tiered Figure 4.12. Main Effects Plot for Factors Effecting 3 Month Densification. 126 gn rs that projects 6) , the affic t at ? = 0.05, but was significant at = 0.1 h level with the onset of hot weather the following year. Superpave design system, with differing binder grades, aggregate properties and Ndesi levels generally accounts for the effect of varying traffic. Fine mixes appear to densify slightly more than coarse mixes. The most interesting effect may be that of month of construction. The numerical month is shown on the x-axis, e.g. April = 4. It appea projects constructed between April (4) and June (6) densified the most, approximately 1 percent more than projects constructed in July (7) and August (8). The fact that constructed in April (4) densified slightly less than the projects constructed in June ( again most likely illustrates the effect of binder aging since the projects constructed in April (4) would have aged slightly before the hottest summer weather. As expected projects constructed in September (9) and October (10) appear to have densified approximately 1 to 2 percent less than the projects constructed in mid-summer. The ANOVA was re-run using the amount of densification after 2 years of tr as the response variable. High PG bump (p = 0.007) was still significant at ? = 0.05. Month of construction (p = 0.068) was not significan ? 0. Figure 4.13 illustrates the fitted means of the effect. This indicates that mont of construction has a strong influence on the long-term densification of a project with approximately a 2 percent change in densification between pavements constructed in May (5) as compared to pavements constructed in October (10). This emphasizes the need to obtain good compaction during late season paving. Compaction requirements cannot be waived with the assumption that the pavement will densify to an acceptable Month of Construction M e an o f 2 - Year D e n s i f i c at i o n , % G m m 10987654 4.5 4.0 3.5 3.0 2.5 2.0 Main Effects Plot (fitted means) for 2-Year Densification ar Densification. Data from the 2000 NCAT Test Track was analyzed in addition to the data from the field projects. The NCAT Test Track offered a unique opportunity to study pavement densification and it?s relationship to the number of design gyrations, since all of the sections receive the same traffic, have the same base and subgrade support and are exposed to the same climatic conditions. Thirty-two of the test track sections were designed using Superpave and are included in the following analysis. The 32 sections represent a range of aggregate types, NMAS, and gradations. One of the objectives of the work at the track was to evaluate densification of HMA. Cores, for evaluating densification, were taken at various traffic levels from the left wheel path of the last 25 feet of each section. When the test track was constructed, paving was carried past the end of the section, and the pavement cut back prior to Figure 4.13. Main Effect Plot for Month of Construction on 2-Ye 127 128 representa . Initi affic in ber h t operation, three trucks w ti ovember 20 a ully ented (four trucks) in February of 2001. For the first three months, cores were n a mo s and later quarterly. The co wed in heir res tive layers and the bulk specific gravity of yer det sing A HTO T 6. D of sam s havin eate 2 t water n was determined using the Corelok device. In- air alcula e construction m um specific gravity values. Figure 4.14 the a rack pavement de ty as ction ALs fo e Su ve s thro pletion of 10 million ESA n Dec er 200 e fig constr on dens s we ghtly lo r for the PG 76- laye d to th ther lay For the PG -22 and 76- s, the n dens s were less for the upper lift. A second-order mial was fit to the data for each binder grade/lift combination. The to indi distinc tes of densification for each binder g lift atio ime af stru n and perature (season). There appears n in f the m betwee e first third points t in ber er of 2000, respectively. The average pavement density appears tinue rom December 2 (thir a point ough O er 2 oint ately 4.5 million ESALs). There is little increase pave bet r 2001 June 2 2 (data point at ap ximately 7.5 m constructing the next section. In this manner, the last 25 feet of the section should be tive mix ally, tr ere opera began onal in N Septem 2000 wit 00 and tr only one ffic was f ruck in implem taken o nthly basi res are sa to t pec each la ermined u AS -16 ensity ple g gr r than percen absorptio place voids were c ted using th axim shows verage test t nsi a fun of ES r th perpa section ugh the com Ls i emb 2. Th ure indicates that the initial ucti itie re sli we 22 surface rs as oppose e o ers. both 67 PG 22 section constructio itie polyno data seems cate t ra rade/ combin n related to t ter con ctio tem to be a itial seating o ix n th and data aken Septem and Decemb to con to increase f 000 d dat ) thr ctob 001 (data p at approxim in ment density ween Octobe and 00 pro illion NCAT Test Track Average Densification PG 67-22 Upper Lift R 2 = 0.98 PG 67-22 Lower Lift R = 0.96 95 96 97 98 y, 2 PG 76-22 Lower Lift R = 0.98 t t % Gmm 2 PG 76-22 Upper Lift R 2 = 0.98 93 94 Pavemen Densi 90 91 92 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 7.E+06 8.E+06 9.E+06 1.E+07 ESALs PG 67-22 Lower Lift PG 76-22 Lower Lift PG 67-22 Upper Lift PG 76-22 Upper Lift Winter 2001-2002 Figure 4.14. Average Test Track Pavement Densification (78). hich the exception of the ster. 22 not densify as fast as the PG 67-22 surface lift. The difference in density was approximately ESALs). In fact, the average density for all but the PG 67-22 upper lift sections appears to decrease in March 2002 (data point at approximately 6.5 million ESALs). The change in density during the summer of 2002 (7.5 to 8.5 million ESALs) is similar to that w occurred during the summer of 2001 (3.0 to 4.5 million ESALs). A slight decrease in density was observed between September and December 2002 with PG 67-22 upper lifts, which increased slightly. There appears to be a significant difference in the rate of densification based on binder grade. As expected, the sections with the softer binder, PG 67-22, densified fa This was true for both the upper and lower lifts. Further, it appears that for the PG 67- sections, the lower lift, which was 50 mm below the surface of the pavement, did 129 130 initial density. Recall that Blankenship (64) did not find a relationship between traffic and pavement densification for layers deeper than 100 mm from the pavement surface. Based on the reduced vertical pressure calculated using Boussinesq theory, Brown and Buchanan (2) recommended Ndesign be reduced by 28 percent or approximately one gyration level for layers deeper than 100 mm from the pavement surface. Brown et al. (78) also note that permanent deformation (and densification) essentially stopped when the air temperature was less than 28 C. Important findings from the densification of the 2000 NCAT Test Track related to this study include (78): 1. Modified binders (2 High PG bump) rutted approximately 60 percent less than unmodified (0.5 High PG Bump) based on an average rut depth after 10 million ESALs of 1.7 mm for the modified mixes and 4.1 mm for the unmodified mixes. 2. one percent from approximately 4 through 10 million ESALs. The difference was not apparent prior to 4 million ESALs because the lower lifts were constructed at a higher Densification was reduced by 25 percent for the surface mixes containing modified binders with an average reduction in air voids of 4.1 percent for the modified mixes and 5.6 percent for the unmodified mixes. The densification of pavement layers 50 mm from the pavement surface was approximately 1 percent less than for surface layers. 4.2.4 Determination of Ndesign to Match Ultimate In-Place Density Three different analyses were performed to relate Ndesign to the ultimate in-place density. Each of these analyses will be described in the following section. First, regressions were performed between the accumulated traffic after two years and the 131 en the accumulated ESALs at each of the sampling intervals (3-months, 6-months, 1-year, 2- rs and 4-years) and the predicted gyrations to match the in-place density at each of those intervals. Third, models were developed to predict Ndesign, which accounted for as-constructed density, high temperature PG grade and traffic. In addition, the ultimate in-place density was compared to the density at the agency specified Ndesign. The number of gyrations necessary to obtain the in-place density after two years of traffic or ultimate density was determined by performing a linear regression between the estimated sample density at a given number of gyrations and the Log gyrations. This was done both for the average densities and pill heights for a project as well as the average density and sample height for each sample within a project (average of 3 SGC pills). The pill height and density at 8, 25, 50, 75, 100, 125 and 160 gyrations were used for the regression to determine the slope and offset. The heights and pill densities from 8, 25, 50, 75, and 100 gyrations were from the SGC pills compacted to 100 gyrations; the pill heights and densities for 125 and 160 gyrations were from the SGC pills compacted to 160 gyrations. It should be noted that the SGC pill densities at 100 and 160 gyrations were measured, but the other pill densities were estimated using Equation 6. References (69, 71, 76) discuss the errors in back calculation of sample density. Due to the scope of the project, back calculation was unavoidable. Once the slope and offset were determined, the number of gyrations to match the ultimate density could be calculated. This was done for both the Pine and Troxler SGCs. Figure 4.15 shows a plot of the average (for each project) Ndesign to match the 2-year in-place density for each SGC predicted Ndesign values. The data were subdivided and potential outliers examined in an attempt to improve the relationship. Second, regressions were performed betwe yea 132 nd - year ESALs and Log predicted gyrations. There appear to be a number of potential outliers. All of the potential outliers are 9.5 mm NMAS mixes and occurred with the Troxler compactor. It also appears that the predicted gyrations for the Troxler compactor are approximately 20 gyrations higher than the predicted gyrations for the Pine compactor. versus estimates of the accumulated traffic after 2-years. The figure is shown with an arithmetic scale, to better show the difference in predicted gyrations between the Pine a Troxler SGCs. The best fit line in the figure is a power model which would produce a straight line on a Log-Log plot. The R 2 values indicate a weak correlation between Log 2 R 2 80 Ma t c = 0.38 R 2 = 0.22 100 120 140 160 0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 12,000,000 c te ra o h Fie l d De ns ity a t Tw o-Ye a r s 0 20 40 60 Cumulative ESALs at Two Years P r e d i d Gy tions t Pine Troxler Power (Pine) Power (Troxler) Potential Outliers Figure 4.15. Predicted Gyrations to Match Two-Year Density. Significant efforts have been made to study the differences in sample density produced by different models and units of gyratory compactors. One influencing factor 133 on posed a DIA of 1.16 ? 0.03 degrees (48). In a study conducted for Alabama DOT, Prowell et al. (49) determined that a change of DIA of 0.1 degrees ill result in a change of 0.01 G mb units as shown in Figure 2.13. Dalton (44) found a imilar f adjusted to an internal angle of 1.16 degrees falls along the line of equality. The best fit line for the original data is shown for comparison. The data in Table 4.9 are sorted by the 20 year design traffic. Lines have been added to the table to separate between the current design traffic levels. In Figure 4.16 and Table 4.9, there appear to be a few potential outliers in the adjusted data, specifically the Pine results for IL-3 and the Troxler results KY-2 and MI-1. The two Troxler points also appeared to be potential outliers in Figure 4.15. One tool for evaluating potential outliers in a relationship is to look at the standardized residual. The standardized residual that has been identified is the dynamic internal angle (DIA) of gyration. The internal angle of gyration can be measured using a device called the dynamic angle verificati kit (DAVK). FHWA pro w s relationship with a change of DIA of 0.1 degrees resulting in a change of 0.014 G mb units. After the completion of this Alabama DOT study, the DIA of the Pine compactor was measured as 1.23 degrees as part of the Alabama DOT study. The DIA o the Troxler compactor was not measured at that time due to a problem with the electronics but was later measured as 1.02 degrees. Using the first relationship, the compacted sample densities from both compactors were adjusted to that which would have been produced if both compactors had been set to a DIA of 1.16 degrees. The predicted gyrations to match the in-place density after two-years of traffic were then recalculated and are summarized in Figure 4.16. As shown in Figure 4.16, the best fit line for the predicted gyrations to match the in-place density from both compactors y = 0.9672x + 2.4663 R 2 = 0.81 y = 1.1005x + 10.837 R 2 = 0.75 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Predicted Gyrations at 2 Years for Pine Compactor Pr ed i c t e d G y ra t i on s at Tw o Ye a r s f o r Tro x l e r Com pactor 1.16 Raw Linear (1.16) Linear (Raw) Line of Equality Figure 4.16. Comparison of Predicted Gyrations to match In-Place Density after Two-Years with and without Correction for DIA. are root of the idered outliers. The standardized residuals for IL-3, KY-2, and ted to ons to s the difference between the observed and the fit values divided by the squi mean square error (MSE). Montgomery (85) states that standardized residuals which exceed ? 3.0 may be cons 134 MI-1 were -2.44, 2.45, and 2.57, respectively; this indicates that they should not be removed as outliers. The other three potential outliers in Figure 4.15, FL-1, MI-2 and MI-3, have standardized residuals of 1.58, 1.09, and 1.33, respectively, when correc a DIA of 1.16 degrees in Figure 4.16. Research by Moseley et al. (86) indicated that the measured DIA is affected by the HMA mixture. Nova Scotia granite, the same as that used in project FL-1, produced the largest differences between compactors; 9.5 mm NMAS mixes also showed larger differences. Observation of Table 4.9 indicates that very few of the predicted gyrati match the in-place density after 2 years exceed the currently specified Ndesign values. 135 ears Average Predicted Gyrations to Match 2 Year Density TABLE 4.9 Original and Adjusted Gyrations to Match In-Place Density at 2 Y Project 20 Tear Traffic, 1.23 1.16 Std. 1.02 1.16 Std. Design ESALs Pine Degrees Pine Degrees Pine Troxler Degrees Troxler Degrees Troxler KY-1 53,706 11 12 1.3 16 14 1.2 KY-3 84,028 34 40 17.4 54 47 22.9 AL-6 143,958 18 20 2.1 26 21 1.3 NE-3 365,719 46 53 10.6 56 44 13.6 NE-1 383,385 47 65 53.3 57 52 39.6 CO-3 523,624 63 69 11.0 77 66 7.4 CO-4 720,911 36 40 11.7 49 42 8.7 CO-1 756,789 62 72 21.7 88 75 18.9 UT-1 771,982 26 28 7.5 36 31 8.8 FL-1 811,658 87 97 14.7 138 115 19 CO-2 1,017,593 44 50 13.9 59 50 13.7 .4 CO-5 1,017,593 37 42 13.5 56 49 15.4 MI-2 1,250,146 74 84 27.4 109 96 32.8 NE-2 1,450,960 69 78 15.1 82 68 13.2 MI-3 1,515,200 86 96 2.5 137 111 5.2 AL-5 1,809,675 25 59 9.2 36 55 9.1 IN-1 1,850,992 47 51 5.0 74 64 5.4 TN-1 3,490,393 33 37 10.1 34 29 10.2 AL-2 3,610,001 38 42 16.5 51 47 20.9 AL-4 4,899,406 59 66 3.0 86 69 4.9 AL-1 6,748,142 54 59 9.2 62 55 9.1 GA-1 8,803,521 47 53 10.7 59 48 5.4 AL-3 8,861,352 31 34 0.5 39 33 1.1 KS-1 10,075,962 50 58 21.6 65 57 20.4 KY-2 12,438,605 77 88 43.9 124 116 57.5 MO-2 12,517,675 68 74 3.6 77 67 6.9 WI-1 14,614,748 58 64 4.9 86 73 9.1 MI-1 15,966,398 91 97 8.9 145 126 15.6 NE-4 20,084,248 83 92 3.0 104 85 5.4 IL-1 26,285,917 73 78 7.2 79 85 10.6 MO-1 27,546,007 93 99 13.0 IL-3 44,466,336 102 109 10.6 96 85 4.9 91 80 6.6 IN-2 45,150,555 54 59 9.1 84 71 9.1 15.0 13.9 4.8 78 69 4.4 NC-1 73,918,507 44 62 16.0 73 60 14.7 AR-2 91,370,805 40 43 5.6 48 42 7.3 AR-4 97,890,077 110 120 3.1 100 111 9.4 AR-3 170,842,507 88 96 14.3 94 86 11.0 IL-2 46,344,297 70 74 17.6 65 53 AR-1 48,726,562 65 72 16.4 81 71 MO-3 53,683,941 68 72 136 The Pine and Troxler results for FL-1 (97 and 115, respectively) exceed 75 gyrations in the > 0.3 to < 3 million ESALs category. The Troxler results for MI-3 (111), KY-2 (116), and MI-1 (126) all exceed 100 gyrations in the >3 to < 30 million ESALs category. The higher numbers for the Troxler compactor may be partially attributed to error in the correction to a DIA of 1.16 degrees. It is expected that if DIA of the Troxler compactor used in this study were measured with the DAVK using these mixes, the measured DIA would be less than the DIA of 1.02 degrees measured in the Alabama DOT study. Figure 4.17 shows the predicted gyrations to match the two-year density, corrected to a DIA of 1.16 degrees, versus the two-year ESALs. Comparison of Figure 4.17 to Figure 4.15 (showing the uncorrected gyration data) indicates that correction of the gyratory data to a common DIA produces similar relationships between two-year y = 10.6 140 6x 0.1312 R 2 = 0 3 y = 8.674x 0.146 R 2 = 13 2 8 100 2-Year ESALs G y r a ti o n to M a tch I n -P l a ce Den s i t y .304 0.37 Potential Outliers 40 60 0 120 0 0 0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 12,000,000 Pine Troxler Power (Troxler) Power (Pine) Potential Outlier Figure 4.17. Predicted Gyrations to Match Two-Year Density Corrected to a DIA of 1.16 Degrees. 137 the additional point, AR-2, appears to be a potential outlier having a low number of predicted atch both the 2-year and 4-year in- re was ESALs and predicted gyrations for the two SGCs, but does not significantly improve R 2 . The same five points discussed previously appear to be potential outliers. An gyrations (43) for a high 2-year traffic level (6.6 million ESALs). Figure 4.18 shows the predicted gyrations to m place densities versus the 20-year design ESALs. Previously, it was shown that the no statistical difference between the 2-year and 4-year in-place density. Figure 4.18 shows a slight increase in predicted gyrations to match the 2-year and 4-year in-place densities for both the Pine and Troxler compactors. However, this appears to be somewhat driven by project AR-4. The in-place density for project AR-4 increased by 0.2 percent between 2-years and 4-years. This resulted in an approximately 9 gyration y = 8.346x 0.126 R 2 = 0.3213 y = 6.5889x 0.1418 R 2 = 0.3893 y = 10.378x 0.1126 R 2 = 0.2572 y = 7.866x 0.131 R 2 = 0.3194 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 c h I - P La ens y 0 2 0,0 00, 0 00 4 0, 000,000 6 0 ,000,000 8 0,0 00, 00 0 1 00, 000 , 00 0 1 20, 000 , 00 0 1 40,0 00 ,0 0 0 1 60, 000 , 00 0 1 80, 000 , 00 0 20-Year Design ESALs Pr edict ed Gy rat i ons to Mat n ce D it Pine 2-Year Pine 4-Year Troxler 2-Year Troxler 4-Year Power (Pine 2-Year) Power (Pine 4-Year) Power (Troxler 2-Year) Power (Troxler 4-Year) Pine 2-Year Pine 4-Year Troxler 2-Year Pine 4-Year AR-4 Pine Troxler Figure 4.18. Predicted Gyrations to Match In-Place Density Corrected to a DIA of 1.16 Degrees. 138 increase between 2- and 4-years. The slight increase in R 2 for the 2- and 4-year relationships is most likely due to missing 4-year data, particularly FL-1. Another way to evaluate whether or not the current Ndesign values are correct is to compare the laboratory air voids at the Ndesign specified by the agency with the in- place density after 2 years of traffic or ultimate density similar to Figure 2.19 (73). Figure 4.19 shows the air voids at Ndesign (1.16 degrees) versus the 2 year in-place air- voids for each of the samples within a project. As expected based on the data presented so far, there is a great deal of scatter in the data. However, the relationship is significant at ? = 0.05. Based on the regression line, at a void level at Ndesign of 4-percent the average in-place air voids are 5.5 percent, or 1.5 percent higher than design. Only a few points fall below the line of equality. This indicates that the pavements have not 0 2 4 12 6 i d s a 8 In - P l a ce A i r Vo fte % 10 r 2 Ye a r s, 1.5% Line of Equality Y = 2.43 82 X S = 1.43 q = 28.3% + 0.7 R-S 024681012 Air Voids at Agency Specified Ndesign, % Analysis riance Source DF MS F P Regres 1 8 7.493 42.64 0.000 Residua 108 2 2.052 Total 109 309. Figure 4.19. In-Place 2 Year versus Agency Specified Ndesign Air Voids. of Va sion SS 7.493 8 l Error 21.609 102 139 y to ere lly recorded and used in the back-calculation. The data have been adjusted to an internal angle of 1.16 degrees. he internal angles of gyration for the two compactors used during the construction of the pper (97 7 gyration difference between the predicted gyrations to match the upper lifts of PG 67-22 densified to their design levels. It further suggests that the Ndesign levels may be too high. By comparison, Harmelink and Aschenbrener (73) found a difference of 1.2 percent after 5 to 6 years based on 22 projects, again indicating that the design levels were too high. Similar to the 40 field projects, the numbers of gyrations to match field densit were back-calculated for the 28 Superpave sections at the 2000 NCAT Test Track. Two Troxler Model 4141 SGC, the same Troxler model used in the field study, were used compact the SGC samples at the 2000 NCAT Test Track. Three replicate samples were compacted for each sublot. The samples were compacted to the same N design level used in the mix design, generally 100 gyrations. The bulk specific gravities of the samples w determined with AASHTO T166. All of the heights were digita T 2000 NCAT Test Track were not known and could not be measured since these compactors were no longer operational. Therefore the average angle, 1.02 degrees, determined for that Troxler model in a previous study was used when adjusting the data to a DIA of 1.16(49). Figure 4.20 shows the average number of gyrations to match the in-place density versus ESALs for a given group of Test Track sections. The data are subdivided by binder grade (PG 67-22 or PG 76-22) and lift (upper surface lift or lower lift 50 mm deep). Second order polynomials provided good fits to the data. On average, there was a 25 gyration difference between predicted gyrations to match the u gyrations) and lower (72 gyrations) PG 67-22 lifts at 10 million ESALs and there is a 3 PG 67-22 Upper Lift R 2 = 0.96 PG 67-22 Lower Lift R 2 = 0.88 PG 76-22 Lower Lift R 2 = 0.90 PG 76-22 Upper Lift R 2 = 0.94 0 10 20 30 40 50 60 70 80 90 100 110 Winter 2001-2002 0 1 000 0 00 2 00 00 00 3 000 0 00 4 0 000 0 0 5 000 0 00 6 0 000 0 0 7 000 0 00 8 0000 00 9 000 0 00 1 0000 000 ESALs Des i gn Gy rations PG 67-22 Lower Lift PG 76-22 Lower Lift PG 67-22 Upper Lift PG 76-22 Upper Lift Figure 4.20. Average Gyrations to Match 2000 NCAT Test Track Density. (97 gyrations) and PG 76-22 (60 gyrations) at 10 million ESALs. As noted previously, no densification occurred during the winter of 2001-2002. Although the relationship is a great deal -22 sections. This is evidenced by the R 2 = 0.63 for the PG 67-22 mixes and R 2 = 0.18 for relationship could be found from which to predict the appropriate Ndesign levels to between the average predicted gyrations and applied traffic is strong, there of scatter in the data. Figure 4.21 presents the actual data for the PG 67-22 and PG 76 upper lifts where each point represents the number of gyrations to match the in-place density for a given section at a given number of ESALs. It is apparent from Figure 4.21 that the scatter in the data is much larger for the PG 76-22 sections than for the PG 67-22 the PG 76-22 mixes. It is possible that if the field data were similarly subdivided, a better match ultimate density. 140 R = 0.18 R 2 = 0.63 20 40 60 70 80 90 110 120 i c t e d Gy r a ti s t o Matc h eld De n s it y 2 0 1 30 50 100 0000 000 000 Pr e d on Fi PG 67-22 0 0 1 000000 2 000000 3 00 4 000000 5 000000 6 000 7 000000 8 0 00000 9 000 1 0000000 ESALs PG 67-22 Upper Lift PG 76-22 Upper Lift The predicted gyrations, corrected to an internal angle of gyration of 1.16 degrees to match the two-year in-place density from the NCHRP 9-9 (1) field projects, excluding the nine projects which used PG 76-22, are shown in Figure 4.22. It is evident fro figure that there is still a great deal of scatter in the data. Three projects with a high number of predicted gyrations for a low design traffic level are CO-1, MI-2, an PG 76-22 Figure 4.21. Predicted Gyrations to Match 2000 NCAT Test Track Density. , m the d MI-3. All three of the projects were constructed with PG 58-28 binder and were constructed with crushed gravel aggregate. Project FL-1 was constructed to 91.8 percent Gmm and d to 95.2 percent Gmm after two years. Nothing appears to be unusual about the re predicted to match the two-year c volume. The laboratory voids for FL-1 were high with densifie densification; however, a high number of gyrations we density for a relatively low traffi 141 R 2 = 0.4117 140 FL-1 120 nsit y MI-3 142 R 2 = 0.3146 0 20 40 60 Pr ed Gy rat i ons to Mat c h 2 - 80 100 0 1 0, 2 0, 3 0 4 5 0, 6 0, , 0 0 0 edict Ye MI-2 ar De CO-1 Pine Non PG 76 Troxler Non PG 76 000,000 000,000 ,000,000 0,00 0, 000 000,000 0 0 0 20 Year Design Traffic, ESALs the Pine lting points for each project. The data for CO-1, MI-2, MI-3 and FL-1 were limina - ts Figure 4.22. Predicted Gyrations Excluding Projects Using PG 76-22. air voids at the agency specified Ndesign of 5.1 and 5.6 percent, respectively for and Troxler compactors. A regression was performed using Log 20-year ESALs as a predictor for Log gyrations. The average Pine and Troxler results at 1.16 degrees were combined resu in two data e ted from the data set. The R 2 = 0.52 indicates a weak correlation between Log 20 year ESALs and Log predicted gyrations. However, the Troxler results for MI-1 were indicated as a possible outlier with a standardized residual of 3.41. The Troxler resul for MI-1 were removed from the data set and the regression re-run. The resulting R 2 (0.57) still indicates a weak correlation, but improved. Figure 4.23 shows the standardized residuals versus the fitted value for the regression. The residuals appear to 143 iction interval for the regression is also shown. Using t the currently sp with the 80 percent confide pre in Table 4.9 is projects constr not include the ted with PG 76-22. From Table 4.10 it can be seen that the hig de o atches the currently sp used the predic a from the 2000 NCAT Test Tr and s should be adeq ic for the 0 N 4.24 it can be s less than appro Ndesign for be be well distributed. Figure 4.24 shows a plot of the regression with the 80 percent confidence interval. The regression was used to predict fitted values for the currently specified traffic levels. The 80 percent pred he regression shown in Figure 4.24, the number of gyrations for each of ecified Superpave traffic levels was calculated along nce diction interval (Table 4.10). The 80 th percentile, calculated using the data shown for comparison. The data for the 80 th percentile includes the ucted with PG 76-22, while the predicted values from the regression does projects construc h si f the interval for the 80 percent prediction interval approximately m ecified gyration levels (4). However, the original Ndesign experiment ted value with 50 percent confidence (64). The dat ack the 80 th percentile data support the fact that an Ndesign of 100 gyration uate for very high traffic levels (Figure 4.20). The 20-year design traff 200 CAT Test Track would be in excess of 100 million ESALs. From Figure een that the predicted gyrations change very rapidly at design traffic levels ximately 3 million ESALs. Caution is required when recommending tween 0.3 and 3 million design ESALs. Fitted Value St 144 a n da r d i z e d R e s i du a l 2.01.91.81.71.61.51.4 2 1 0 -1 -3 -2 Gyratio Figure 4.23. Standardized Residuals versus Fitted Mean for Log Predicted ns versus Log 20-Year ESALs. logten(SGC Average 1 20 Year Design Traffic, ESALS P e d i c d G y a ti o to a tc h r te r n s M U a e l ti m te D n s i t y 6 0 0 0 0 0 0 0 5 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 00 140 S 0.132262 R-Sq 57.2% 2 0 0 0 0 0 1 0 0 0 0 0 120 100 R-Sq(adj) 56.4% Regression 80% CI 80% PI .16) = 0.5993 + 0.1729 logten(20 Year ESALS) 80 60 40 20 0 Figure 4.24. Predicted Gyrations versus 20 Year Design Traffic without PG 76-22 Data. 145 80 % Prediction 80th Percentile TABLE 4.10 Predicted Gyrations to Match Ultimate Density Interval 20 Year Design Current Predicted Low High Pine Troxler Avg. ESAL Ndesign Ndesign 300,000 50 35 23 53 32 43 37 1,000,000 75 43 29 65 71 73 72 3,000,000 100 52 35 78 83 90 87 10,000,000 100 65 43 96 59 55 57 30,000,000 125 78 52 117 95 104 100 100,000,000 125 96 64 145 101 82 92 Figure 4.25 shows the relationship between 20-year design ESALs and the predicted gyrations to match the 2-year density for the projects constructed with PG 76- 22. Although a best fit line is shown in the figure, there is no relationship between the 20-year design ESALs and the predicted gyrations for the projects constructed with PG 76-22. A poor relationship (R 2 = 0.18) was also observed for the data from the 2000 NCAT Test Track (Figure 4.21). This indicates that for the modified binders there was y = 24.869x 0.0575 R 2 = 0.0697 y = 29.2 R 2 = 0. 0 20 40 60 80 100 0 0 4 6 0 , 0 0 0 , 0 0 0 , 0 0 0 0 0 1 4 0 , 0 0 0 0 , 0 0 0 1 8 20 Yea Ls tch 2 - Y ear Den 13x 0.047 0405 12 140 s ity 2 0 , 0 0 0 , 0 0 0 0 , 0 0 0 , 0 0 0 0 0 , 0 0 0 0 8 0 , 0 0 1 0 0 , 0 1 2 0 , 0 0 0 , 0 0 , 0 0 0 1 6 0 , 0 0 , 0 0 0 , 0 0 0 r Design Traffic, ESA Predicte d Gyration s to M a Pine Troxler Power (Pine) P Troxler) Figure 4.25. Predicted Gyrations for Projects with PG76-22. ower ( 146 no co ti When the Ndesign table was originally developed, regression analysis was perfo b gyrations required to match the in-place pavement density after more than 12 years of traffic ). The analyses for the NCHRP 9-9 (1) field section presented thus far have been based solely on the number of gyrations to match the ultimate pavement density (2- year or 4-year). Figures 4.20 and 4.21 presented the predicted gyrations to match in- place it sents a log-log plot of predicted gyrations versus accumulated traffic for all of the NCHRP 9-9 r the io As expected, there is considerable scatter in the data as evidenced by the low rrela on between change in density and traffic. rmed etween gyrations determined to match the as-constructed density and the (64 dens y for the 2000 NCAT Test Track as traffic accumulated. Figure 4.26 pre (1) pre fie dict ld sec ns. tions. The gyratory data corrected to a DIA of 1.16 degrees were used fo 1 y = 8.5477x 0.1433 R 2 = 0.3727 y = 9. 2 603x 0.1347 R = 0.3406 10 1 10 100 1,000 10,000 100,000 1,000,000 10,000,000 100,000,000 Log Cumulative Traffic, ESALs 1 100 000 L o g P r e d i c t e d G y r a t i o n s t o M a t c h I n - p l a c e D e n s i t y Pine Troxler Power (Pine) Po nsi w ty er ( for Tro al xler l P ) ost 4.26 uction Sampling Periods. Fig Co ure nstr . Predicted Gyrations to Match In-Place De - 147 R 2 values. The regression line for the Pine and Troxler da imately identical. The equatio ratio ilar to Table 4.10 and are presented in Table 4.11. This method of analysis produces slightly higher predic g ie TABLE 4.11 Predicted Gyrations to Match In-Place Density ta are approx ns for the best fit line were used to predict gy n levels sim ted yration levels, close to those currently specif d. 20-Year Design ESAL Troxler Pine 300,000 52 52 1,000,000 62 62 3,000,000 73 74 30,000,000 101 100,000,000 120 126 10,000,000 86 89 105 con con tha as s Reg R One c ted d ed i oncern about the predictions is the high percentage of projects with low as- stru str t the c ensity. Figure 4.6 indicates that 55 percent of the projects had as- data suggested w ate (2-year) density hown in Figure 4.27 for both the field projects and the 2000 NCAT Test Track. re n - densities indicated 0.65 and R 2 = 0.54 for the field projects and 2000 NCAT Test Track, respectively. s in ts 0 NCAT Test ck is somewhat expected based on the accelerated traffic loading at the 2000 NCAT t T . Since a higher as-constructed density would her ultimate si i h ttempt was made odel pavement densification to predict in-place density. It was felt that this model ossi d .g. 92 or 93 as-constructed den 96 percent ultimate de uct n-place densities less than 92 percent. Examination of the re as a strong trend between as-constructed density and ultim ssio analyses between the as-constructed and two year in-place 2 = The Tra Tes den to m cou per hift the regression lines between the field projec and the 200 rack result in a hig ty, th s could affect the predicted Ndesign levels. T erefore, an a ld p cent bly be used to pre sit dic y a t N nd design with ideal fiel co nsi ndi ty. tions, e y = 0.7147x + 29.144 R 2 = 0.65 y = 0.815x + 21.569 R 2 = 0.54 85 8 .0 85 7.0 89.0 91.0 9 95.0 97.0 9 89.0 91.0 93.0 95.0 97.0 99.0 As-Constructed Density, %Gmm I n - P l a c e De ns i t y af ter 2 Ye ar s , %G m m 3.0 9.0 .0 87.0 NC -9 (1)HRP 9 2000 NCAT Test Track Linear (NCHRP 9-9 (1)) Linear (2000 NCAT Test Track) Line of Equality tween As-Constructed and 2-Year (Ultimate) Density. pected to affect pavement densification. viously, high temperature PG bump and month of construction were shown to be if t f t pavement densification. Brown and Cross (60) suggested the Log of accumulated ESALs divided by the Log of the design compaction effort th se good predictor for in-place density. Based on the literature, as suggested that pavements constructed to a low initial density would tend to densify entually obtain the same ultimate density as pavements constructed to higher ial si ndicates that there is a weak trend of increased sification for projects with lower as-constructed densities, but no trend for projects cept ensities. Therefore, the difference between the laboratory at Ndesign and the as-constructed density was considered as an alternative. Fig Pre sign that (in it w more and ev init den wit den ure 4.27. Relationship be Epps et al. (58) described factors ex ican actors which affec is ca gyrations) was a den ties. Figure 4.28 i h ac sity able construction d 148 149 y = -0.609x + 58.569 y = -0.2158x + 22.668 R 2 = 0.03 2.0 3.0 4.0 5.0 6.0 7.0 93.0 94.0 95.0 96.0 %Gmm r Den s if icat ion (D elt a ), %Gmm R 2 = 0.43 0.0 1.0 87.0 88.0 89.0 90.0 91.0 92.0 As-Constructed Density, Two- Yea Low As-Constructed Density Acceptable As-Constructed Density Linear (Low As-Constructed Density) Line Figure 4.28. Two-Year Densification A number of techniques, such as best subsets and number of iterations were attempted to develop a m density. Variables used to predict 2-year density included: average annual air temperature, NMAS, high PG grade, agen gyrations, month of construction, attempt was made to model pavement densific found. Better results were obtained when predicting pavem models developed is Equation 9: ofMonthDenConstDenYear ar (Acceptable As-Constructed Density) versus As-Constructed Density. step-wise regression, and a odel to predict the 2-year pavement Degree days over 30?C, mean cy specified design 2-year ESALs, and as-constructed density. Initially, an ation, but not even a fair model could be ent density. One of the best PGHighConst ?????= 325.0..771.0.2 Month of construction was entered as the numerical m 078.0. (9) onth of construction, e.g. July = 7. High PG is the high PG binder grade, e.g. 64, 67, 70, or 76. The model has an R 2 = 0.71 150 with a standard error = 0.91 and a Mallow?s C-p statistic of 5.6. All of the variables in the model are significant (? = 0.05). It is generally desirable to have a Mallow?s C-p statistic less than the number of variables in the model. This model only represents a slight improvement over the prediction made with just as-constructed density (Figure 4.27, R 2 = 0.65). Minitab?s best subsets analysis identified a five variable model with a Mallow?s C-p statistic of 4.5. In addition to as-constructed density, month of construction and high temperature PG grade, this model included degree days over 30? C and Log of 2-Year ESALs. Degree days over 30? C was determined for each project from LTPPBind version 2.1 (77). If on a given day the temperature were 35? C, that day would account for five degree days. The reported value is the average yearly cumulative degree days. The data set contained projects with from 0 to 444 degree days over 30? C. Regions in the southwestern U.S. have much higher values for degree days over 30? C. For example, Phoenix, AZ has approximately 1400 degree days over 30? C. Equation 10 presents the second model developed for predicting 2-year (ultimate) density: ESALSYLogCDD MCPGHighACDDensityY 2321.0300041.0 204.0132.0786.061.302 ?+?+ ?????+= (10) where, 2Y Density = in-place density after 2-years of traffic, ACD = as-constructed density, High PG = high temperature PG grade, MC = month of construction (July = 7), 30CDD = degree days over 30? C, and 151 2Y ESALs = accumulated ESALs at 2 years. Equation 10 has an R 2 0.76 and a standard error of 0.88. Degree days is not significant at the 5 percent level but is significant at the 10 percent level. The p-value for Log 2-year ESALs is 0.182, indicating that it is not significant. The fact that accumulated traffic is not strongly related to densification is not completely surprising since the projects were designed with a tiered system where projects with higher traffic levels tended to have more angular aggregates, stiffer binders and higher design gyration levels. The models were then used to recalculate the 2-year density for each project assuming that the as-constructed density was 92 percent (the actual values were used for all of the other variables). The number of gyrations to match the new 2-year density (based on a 92 percent as-constructed density) was calculated for each project. Unfortunately, the resulting predicted gyrations produced even poorer relationships with design traffic than those presented previously. This tends to indicate that the scatter in the predicted gyration versus ESAL data was not due to the range of as-constructed densities. Another source for the scatter in the predicted gyration versus ESAL data might be the fact that the HMA for the different projects were not all produced at 4 percent air voids. A project constructed with higher laboratory air voids would be less likely to densify in the field and a project constructed with low laboratory air voids would be more likely to densify in the field. One way to address this issue would be to look at the field densities as a percent of laboratory density. A model was developed to predict Ndesign as a function of high temperature PG grade and Design ESALs. As-constructed density 152 was normalized to 92 percent Gmm in the model development. The following steps summarize the model development: 1. Express the 2-year in-place density for each project as a percent of Gmb (laboratory density) determined at 100 gyrations for both the Pine and Troxler SGCs normalized to a DIA of 1.16 degrees. 2. Develop a model to predict the 2-year percent of laboratory density similar to Equations 9 and 10. Models were developed to predict laboratory density (% Gmb) as a function of as-constructed density, high temperature PG and ESALs. 3. Develop a matrix of twelve 2-year in-place densities based on as- constructed densities of 92 percent, two high temperature PG grades (64 and 76) and a range of design traffic (Table 4.12). 4. Determine the in-place density (%Gmm) corresponding to each of the predicted laboratory densities (%Gmb) in Table 4.12 for each project. 5. Determine the number of gyrations needed to match each of the in-place densities determined in Step 4. The range of gyrations for each percent of laboratory density determined in Step 3 is relatively small. Essentially this says that the SGC compacted all of the mixes in this study at approximately the same rate. This makes sense sine the SGC is a constant strain compaction device. The average number of gyrations to match each of the percent of laboratory density in Table 4.12 was determined for both the Pine and Troxler SGCs. 153 6. Finally, a model was developed to relate Ndesign back to high temperature PG grade and Log ESALs. As-constructed density dropped out of the model since it was set to 92 percent Gmm in all cases. This was accomplished through the percent of laboratory (Gmb) density described in Steps 1-5. The 2-year in-place density expressed as a percent of the laboratory density determined at 100 gyrations was regressed against the same sent of predictors used previously (Step 2). Equations 11 and 12 present the models developed for the Pine and Troxler compactors, respectively: (11) ESALsYLogHPGACDDensityLabPineYear 219.158.0452.095.53%2 ?+???+= ESALsYLogHPGACDDensityLabTroxlerYear 206.108.0381.034.62%2 ?+???+= (12) where, ACD = as-constructed density, High PG = high temperature PG grade, and 2Y ESALs = accumulated ESALs at 2 years. The R 2 = 0.53 for the Pine model and R 2 = 0.45 for the Troxler model with standard errors of 1.27 and 1.28, respectively. The high PG grade was not significant in either model, with p-values of 0.235 and 0.129 for the Pine and Troxler data, respectively. These variables were selected since the produced reasonable R 2 values for both compactors. Better models were identified for one or the other compactor, but they did not share the same variables. A matrix of variables was developed to examine the effect of determining the predicted gyrations to match a given percentage of laboratory density (Step 3). Table 154 4.12 presents the matrix of variables and the resulting percentages of laboratory density. The in-place density corresponding to each of the percentages of laboratory density shown in Table 4.12 was calculated for each project (Step 4). Then the number of gyrations to match that in-place density was calculated for each project (Step 5). The predicted gyrations to match each of the percentages of laboratory density are shown in Tables 4.13 and 4.14 for the Pine and Troxler compactors, respectively. For the Pine compactor, the predicted gyrations for a given percentage of laboratory density had a low variability with standard deviations ranging from 3.44 to 8.99. The predicted gyrations to match a given percentage of laboratory density for the Troxler compactor also had low variability with standard deviations ranging from 4.83 to 8.98. Thus regardless of the mix, a given percentage of laboratory density (determined at an Ndesign of 100 gyrations) can be achieved with a similar number of gyrations. TABLE 4.12 Matrix of Predicted Percentage of Laboratory Density As- Constructed Density 2-Year ESALs Log 2 Year ESALS Approximate 20-Year ESALs High PG Pine Predicted 2-Year %Gmb (Lab Density) Troxler Predicted 2-Year % Gmb (Lab Density) 92 30,000 4.48 300,000 64 97.2 97.3 92 90,000 4.95 1,000,000 64 97.8 97.8 92 230,501 5.36 3,000,000 64 98.3 98.2 92 920,577 5.96 10,000,000 64 99.0 98.9 92 2,583,607 6.41 30,000,000 64 99.5 99.3 92 6,773,140 6.83 100,000,000 64 100.0 99.8 92 30,000 4.48 300,000 76 96.5 96.4 92 90,000 4.95 1,000,000 76 97.1 96.9 92 230,501 5.36 3,000,000 76 97.6 97.3 92 920,577 5.96 10,000,000 76 98.3 98.0 92 2,583,607 6.41 30,000,000 76 98.8 98.4 92 6,773,140 6.83 100,000,000 76 99.3 98.9 TABLE 4.13 Pine Predicted Gyration to Match Percentage of Lab Density Percent of Lab Density, %Gmb 97.2 97.8 98.4 99.0 99.6 100.2 96.5 97.1 97.7 98.3 98.9 99.5 Project Predicted Gyrations AL-1 50 58 68 79 91 107 42 49 57 66 77 90 AL-2 55 64 72 83 94 108 47 55 62 71 81 93 AL-3 42 51 62 75 91 111 33 41 49 60 72 89 AL-4 34 43 54 69 86 109 26 33 41 53 66 83 AL-5 33 42 53 67 84 108 25 32 40 51 64 82 FL-1 43 52 61 73 86 104 35 42 50 60 70 84 MI-1 52 60 69 80 91 105 44 51 59 68 77 89 MI-2 48 57 66 78 91 107 40 47 55 65 75 89 WI-1 44 53 63 76 90 108 36 43 51 62 73 88 CO-1 38 47 57 71 86 107 30 37 45 56 68 84 CO-2 38 47 57 71 86 106 30 37 45 56 68 84 CO-3 44 53 62 74 87 103 36 43 51 61 71 85 CO-4 47 55 65 77 90 107 38 46 53 63 74 88 CO-5 46 55 64 76 89 106 38 45 53 63 73 87 IN-1 54 62 71 81 92 106 47 53 61 70 79 91 IN-2 40 49 58 71 85 103 32 39 47 57 68 83 KY-1 58 66 75 85 96 109 50 57 64 73 83 94 KY-2 58 66 74 84 94 107 50 57 64 73 82 93 KY-3 42 51 61 74 89 108 34 41 49 60 71 87 AL-6 33 42 54 70 89 115 24 32 40 52 66 86 AR-1 52 61 71 83 96 113 43 51 59 69 80 94 AR-2 52 61 71 83 96 112 44 51 59 69 80 94 AR-3 47 56 67 80 95 114 38 46 54 65 77 93 AR-4 42 50 60 72 85 102 34 41 48 58 69 83 GA-1 34 44 55 71 89 115 26 33 41 53 67 87 155 TABLE 4.13 Pine Predicted Gyration to Match Percentage of Lab Density (Continued) Percent of Lab Density, %Gmb 97.2 97.8 98.4 99.0 99.6 100.2 96.5 97.1 97.7 98.3 98.9 99.5 Project Predicted Gyrations IL-1 56 64 72 82 93 107 48 55 62 71 80 92 IL-2 56 64 73 84 96 111 47 55 62 72 82 94 IL-3 54 62 71 82 93 107 46 53 60 69 79 91 KS-1 43 52 62 75 89 107 35 42 50 60 72 87 MI-3 40 49 59 72 87 107 32 39 47 57 69 85 MO-1 58 67 76 87 98 113 50 57 65 74 84 97 MO-2 54 63 71 82 93 107 47 54 61 70 79 91 MO-3 55 64 73 84 96 110 47 55 62 72 81 94 NC-1 29 39 51 68 88 117 21 28 37 49 64 85 NE-1 30 39 50 65 84 110 22 29 37 48 62 81 NE-2 36 45 56 70 86 108 28 35 43 54 67 83 NE-3 27 36 46 61 78 103 20 26 34 45 57 76 NE-4 37 46 56 70 86 107 29 36 44 55 67 83 TN-1 36 46 57 71 88 111 28 36 44 55 68 86 UT-1 47 56 66 78 91 108 39 46 54 64 75 89 Minimum 27.1 35.7 46.0 60.6 78.0 102.4 19.9 26.2 33.8 44.5 57.3 75.5 Average 44.6 53.4 63.1 75.8 89.8 108.3 36.5 43.7 51.5 61.7 73.0 87.8 Maximum 58.0 66.5 75.5 86.7 98.4 116.6 50.2 57.2 64.7 74.3 84.3 96.8 Std. Dev. 8.97 8.61 7.82 6.30 4.33 3.44 8.95 8.99 8.72 7.95 6.67 4.60 156 TABLE 4.14 Troxler Predicted Gyration to Match Percentage of Lab Density Percent of Lab Density, %Gmb 97.2 97.8 98.4 99.0 99.6 100.2 96.5 97.1 97.7 98.3 98.9 99.5 Project Predicted Gyrations AL-1 52 60 68 78 89 102 41 48 54 62 71 81 AL-2 58 66 73 83 92 104 48 54 60 68 76 85 AL-3 43 52 60 72 84 99 33 39 46 54 63 75 AL-4 36 44 54 67 81 101 25 31 38 47 57 71 AL-5 35 42 51 62 75 91 25 31 37 45 54 66 FL-1 46 54 63 75 87 103 35 41 48 57 66 78 MI-1 53 61 69 79 89 102 43 49 55 63 71 82 MI-2 49 57 65 76 88 103 38 44 51 59 68 80 WI-1 45 54 63 74 87 102 34 41 48 56 66 78 CO-1 42 51 60 72 85 101 32 38 45 53 63 75 CO-2 44 53 62 75 88 106 32 39 46 55 66 79 CO-3 46 54 63 73 84 99 36 42 49 57 65 77 CO-4 48 56 65 76 87 102 37 44 50 59 68 79 CO-5 47 55 63 74 86 100 36 42 49 57 66 78 IN-1 54 62 69 79 89 101 44 50 56 64 72 82 IN-2 41 50 59 71 83 100 31 37 44 52 62 74 KY-1 58 66 73 83 92 104 48 54 60 68 76 85 KY-2 59 66 74 84 93 105 48 55 61 69 77 87 KY-3 42 51 60 71 84 101 32 38 45 53 63 75 AL-6 34 42 52 65 81 101 23 29 36 45 56 70 AR-1 54 63 71 82 93 106 43 50 57 65 74 85 AR-2 56 64 72 83 94 108 44 51 58 66 75 86 AR-3 56 65 74 86 98 113 45 51 59 68 77 89 AR-4 50 59 69 81 95 112 38 45 52 62 72 85 GA-1 36 45 55 68 83 103 26 32 39 48 58 73 157 TABLE 4.14 Troxler Predicted Gyration to Match Percentage of Lab Density (Continued) Percent of Lab Density, %Gmb 97.2 97.8 98.4 99.0 99.6 100.2 96.5 97.1 97.7 98.3 98.9 99.5 Project Predicted Gyrations IL-1 56 63 71 80 89 101 46 52 58 65 73 83 IL-2 59 66 74 84 94 107 48 54 61 69 77 87 IL-3 57 64 72 82 92 104 46 52 59 67 75 85 KS-1 46 55 64 76 88 104 35 42 49 57 67 79 MI-3 41 49 58 70 83 100 30 36 43 51 61 74 MO-1 60 68 76 86 96 109 49 56 62 71 79 89 MO-2 57 65 72 81 91 103 47 53 59 67 75 84 MO-3 58 65 73 83 93 105 47 53 60 67 76 86 NC-1 30 38 46 57 69 85 22 27 32 40 49 60 NE-1 33 41 51 64 79 99 22 28 35 44 54 68 NE-2 39 47 57 70 84 102 28 34 41 50 60 74 NE-3 30 38 47 60 75 95 20 25 32 40 50 64 NE-4 40 48 57 68 81 98 29 35 42 50 60 72 TN-1 37 46 56 68 83 102 27 33 40 49 59 73 UT-1 48 56 65 75 86 100 38 44 50 59 67 78 Minimum 29.5 37.6 45.7 56.6 68.7 85.0 19.9 25.3 31.6 40.0 48.6 60.1 Average 46.9 55.0 63.6 74.7 86.7 102.1 36.3 42.4 49.0 57.5 66.6 78.3 Maximum 60.3 68.2 76.2 86.2 97.7 112.9 49.3 55.8 62.4 70.6 79.0 89.3 Std. Dev. 8.98 8.78 8.30 7.39 6.19 4.83 8.81 8.98 8.96 8.66 8.08 7.05 158 159 Since the gyrations were related to the percentage of laboratory density at 100 gyrations, and since the percentage of laboratory density was related to as-constructed density, high PG grade and ESALs, the data were analyzed to see if a relationship existed between the average predicted gyration and high PG grade and ESALs (Step 6). Since a single target as-constructed density was desired (92 percent), this variable should drop out of the relationship. Higher as-constructed densities would (using Equations 11 or 12) result in higher predicted gyrations. Although this seems counter intuitive from a field compaction standpoint, if a mix was constructed to a higher level of density initially, one would want it to be more resistant to additional densification. Likewise a pavement constructed to a lower as-constructed density would tend to age faster, producing a stiffer mix. Therefore, one would need a mix that would densify more readily to achieve the same ultimate density. Table 4.15 shows the data used to develop the models to predict Ndesign gyration levels from high PG grade and 2-year ESALs. The average gyrations to match a percentage of laboratory density are those shown in Tables 4.13 and 4.14 to meet the percentage of lab density determined for the matrix in Table 4.12. Since the Pine and Troxler number of gyrations to match a percentage of laboratory density at a DIA of 1.16 degrees were so close to each other, they were averaged. Two models were then developed between ESALs, High PG grade and gyrations, one using the 2-year ESALs (Equation 13) and one using the 20-year ESALs (Equation 14). Equation 14 was determined following the same steps as Equation 13 using the 20-year ESALs. 160 TABLE 4.15 Matrix of Gyrations Predicted Gyrations to a Percentage of Lab Density 2-Year ESALs 20-Year ESALs High PG Avg. Pine Gyrations to a Percentage of Lab Density Avg. Troxler Gyrations to a Percentage of Lab Density Average Gyrations to a Percentage of Lab Density Eq.13 Eq.14 30,000 300,000 64 45 47 46 46 46 90,000 1,000,000 64 53 55 54 56 56 230,501 3,000,000 64 63 64 63 64 66 920,577 10,000,000 64 76 75 75 77 76 2,583,607 30,000,000 64 90 87 88 86 86 6,773,140 100,000,000 64 108 102 105 95 96 30,000 300,000 76 37 36 36 31 30 90,000 1,000,000 76 44 42 43 41 41 230,501 3,000,000 76 51 49 50 49 50 920,577 10,000,000 76 62 58 60 62 61 2,583,607 30,000,000 76 73 67 70 71 71 6,773,140 100,000,000 76 88 78 83 80 81 (13) ESALsYearLogHPGNdesign 29.2025.10.33 ?+??= ESALsYearLogHPGNdesign 201.2027.18.16 ?+??= (14) where, Ndesign = the number of design gyrations, HPG = high PG grade, and 2-Year or 20-Year ESALs = the 2-year or 20-year design ESALs for the project. The R 2 for both Equation 13 and Equation 14 is 0.97 with standard errors of 3.66 and 3.54, respectively. Note that the model reduces Ndesign by approximately 15 gyrations for a two grade bump in high PG grade (e.g. 64 to 72). The lowest traffic level is equivalent to the 50 gyrations currently specified in AASHTO R 35 for less than 300,000 ESALs. The predicted Ndesign for unmodified binders for the highest traffic level is approximately 25 gyrations less than currently specified in AASHTO R 35 (125 161 gyrations). Further the predicted gyrations for the unmodified binder (PG 64) approximately match those determined in Table 4.10 (presented previously), but are slightly higher in the 10 to 30 million 20-year ESAL range. 4.2.5 Evaluation of Locking Point The locking point concept was developed by Illinois DOT (75, 76). Since its development, other agencies have altered the definition of the locking point. The original definition is the first instance of three consecutive gyrations having the same sample height immediately preceded by two instances of two consecutive gyrations resulting in the same sample height (locking point 3-2-2). Other values used include: first instance of two consecutive gyrations resulting in the same sample height (locking point 2-1), second instance of two consecutive gyrations resulting in the same sample height (locking point 2-2), the third instance of two consecutive gyrations resulting in the same sample height (locking point 2-3) and One criticism of the locking point was that there was little research to tie the results to a physical quantity in the field. The locking point was determined manually for each of the cases described above. One encouraging aspect of the locking point calculations was that the locking point was approximately the same number of gyrations for the Pine and Troxler SGCs without any adjustments (Figure 4.29). However, the density at a given definition of the locking point was higher for the Pine compactor (Figure 4.30), if the data are not corrected to a DIA of 1.16 degrees. Comparisons were made between the calculated density at the four different definitions of the locking point and as-constructed and two- year in-place density. The 2-1 locking point overestimated the as-constructed density as 162 seen in Figure 4.31. The 3-2-2 locking point appears to provide the best relationship with ultimate density (Figure 4.32). However, the relationship is poor, weaker than that determined using design traffic. Various subdivisions of unmodified and modified binder were attempted, since binder stiffness should not affect the results during compaction. The best relationship (R 2 = 0.47) was determined for the projects with modified binders based on the Troxler densities for the 3-2-2 locking point. However, 3 of the 20 projects, AR-3, AR-4 and IL-2, had missing data which prevented their inclusion. y = 0.9748x + 3.1728 R 2 = 0.98 0 10 20 30 40 50 60 70 80 90 100 110 120 0 1020304050607080901010120 Pine 3-2-2 Locking Point Trox ler 3-2 - 2 Loc king Point Line of Equality Figure 4.29. Comparison between 3-2-2 Pine and Troxler Locking Point. y = 0.743x + 25.343 R 2 = 0.77 93.0 94.0 95.0 96.0 97.0 98.0 99.0 93.0 94.0 95.0 96.0 97.0 98.0 99.0 Troxler Desity at 3-2-2 Locking Point Pin e D ens it y at 3-2-2 L o ck ing Point Figure 4.30. Comparison of Average Pine and Troxler Density at 3-2-2 Locking Point. y = 0.7201x + 23.883 R 2 = 0.2637 y = 0.6905x + 27.214 R 2 = 0.2474 87.0 88.0 89.0 90.0 91.0 92.0 93.0 94.0 95.0 96.0 97.0 87.0 88.0 89.0 90.0 91.0 92.0 93.0 94.0 95.0 96.0 97.0 Locking Point 2-1 Density, % As- C on st ruc t ed Des n it y, % Pine Troxler Linear (Pine) Linear (Troxler) Figure 4.31. 2-1 Locking Point Density versus As-Constructed Density. 163 Pine y = 0.6114x + 36.101 R 2 = 0.21 Troxler y = 0.6561x + 32.323 R 2 = 0.29 91.0 92.0 93.0 94.0 95.0 96.0 97.0 98.0 99.0 91.0 92.0 93.0 94.0 95.0 96.0 97.0 98.0 99.0 3-2-2 Locking Point Density, % 2 - Ye ar I n -Place D ens ity , % Pine Troxler Linear (Pine) Linear (Troxler) Figure 4.32. 3-2-2 Locking Point Density versus 2-Year Density. The use of the 3-2-2 locking point would appear to be a conservative way to estimate the ultimate density of the pavement. One potential concern about the use of the locking point is the lubricating effect of binder content on the number of gyrations determined for the locking point. If the asphalt content selected for the locking point determination is on the dry portion of the VMA curve, then the locking point may be higher, whereas if it is on the wet side it may be lower than or close to the locking point at the optimum asphalt content. An evaluation of the locking point over a range of binder contents is beyond the scope of this study. Also, the locking point appears to be a function of the aggregate type, angularity and gradation and is not related to the design traffic. 164 165 4.2.6 Pavement Condition after Four Years Visual assessments were conducted along with the pavement coring at each coring interval. Rut depths were measured with a six-foot string line. Table 4.16 presents the 4-year rut depth measurements. The maximum observed rutting averaged 6.4 mm. The average rutting observed for all of the projects was 1.7 mm. The Superpave mixes are all very rut resistant. Noticeable raveling was observed on 14 of the projects; 13 projects exhibited cracking; 13 projects had popouts; and 7 projects exhibited moisture damage in either the test layer or the underlying layer. The rut depths from the field projects match the findings of the 2000 NCAT Test Track. Brown et al. (2004) reported an average rut depth after 10 million ESALs in two years of 2.7 mm with a maximum rut depth of 7.4 mm. The two sections with the most rutting, N3 (7.4 mm) and N5 (7.1 mm) were both placed with asphalt contents approximately 0.5 percent above optimum. Brown et al. also noted that sections containing PG 76-22 rutted 60 percent less than sections constructed with unmodified PG 67-22. It should be noted that the majority of the observed ?rutting? was attributed to pavement densification under traffic. 166 TABLE 4.16 Four-Year Rut Depth Measurements Sublot 1 Sublot 2 Sublot 3 Core Location Project 1 2 3 1 2 3 1 2 3 Avg., mm Std.Dev., mm AL-1 2 2 2 2 1 3 - - - 2.0 0.83 AL-2 3 2 2 0 0 2 5 5 6 2.7 2.03 AL-3 AL-4 AL-5 FL-1 MI-1 10 9 9 6 7 7 3 2 4 6.4 2.60 MI-2 2 2 2 1 0 2 - - - 1.3 0.82 WI-1 0 0 0 0 0 0 0 0 0 0.0 0.00 CO-1 3 2 4 5 5 3 7 6 7 4.8 1.77 CO-2 1 2 2 2 3 3 3 3 3 2.6 0.79 CO-3 0 0 0 0 0 0 - - - 0.0 0.00 CO-4 2 1 1 2 2 1 2 2 2 1.5 0.62 CO-5 3 2 2 5 5 5 4 3 3 3.6 0.98 IN-1 2 3 2 0 0 2 3 5 2 2.2 1.53 IN-2 3 2 2 5 3 4 3 2 2 3.0 0.95 KY-1 0 0 0 0 0 0 - - - 0.0 0.00 KY-2 1 2 0 0 0 0 - - - 0.4 0.66 KY-3 1 0 0 0 0 1 0 0 0 0.2 0.37 AL-6 0 0 0 0 0 0 - - - 0.0 0.00 AR-1 2 2 2 2 2 2 2 2 2 1.9 0.40 AR-2 3 2 3 3 3 2 - - - 2.8 0.66 AR-3 3 3 2 2 1 3 - - - 2.2 1.06 AR-4 2 2 2 2 2 2 2 2 2 2.1 0.40 GA-1 1 1 1 1 1 0 2 1 0 0.7 0.48 IL-1 0 1 0 1 1 1 0 0 1 0.4 0.42 IL-2 1 3 2 3 3 3 3 2 2 2.6 0.79 IL-3 0 0 0 0 0 1 1 2 2 0.5 0.69 KS-1 1 0 1 1 2 2 2 1 1 1.0 0.53 MI-3 0 0 0 0 0 0 0 0 0 0.0 0.00 MO-1 1 2 1 2 1 3 4 1 2 1.9 1.19 MO-2 2 2 2 1 1 2 3 2 2 1.7 0.74 MO-3 0 0 0 1 0 0 0 0 0 0.1 0.26 NC-1 6 2 2 1 2 2 1 2 1 2.0 1.73 NE-1 2 5 4 2 2 2 - - - 2.5 1.46 NE-2 1 1 1 2 2 2 2 2 2 1.5 0.62 NE-3 2 2 2 2 2 2 2 2 2 2.2 0.35 NE-4 1 1 2 2 2 4 2 2 1.8 1.02 TN-1 2 2 2 3 2 3 2 2 1 2.0 0.80 UT-1 0 0 0 0 0 0 0 0 0 0.0 0.00 167 4.2.7 Evaluation of Ninitial The densities at Ninitial, corrected to a DIA of 1.16 degrees, are shown in Table 4.17. Table 4.17 is sorted by 20-year traffic. AASHTO M 323-04 specifies that the density at Ninitial shall be less than 91.5 percent for 20-year traffic levels less than 300,000 ESALs, less than 90.5 percent for traffic levels between 300,000 and 3,000,000 ESALs, and less than 89.0 percent for traffic levels greater than 3,000,000 ESALs. Based on Table 4.12, none of the samples from projects with design traffic less than 300,000 ESALs fail Ninitial, 36 percent of the samples with design traffic levels between 300,000 and 3,000,000 ESALs fail Ninitial, and 26 percent of the samples with design traffic levels greater than 3,000,000 ESALs fail Ninitial. Failures occur in 11 of the 40 projects. The mixes are fine-graded for 9 of the 11 projects that fail Ninitial. Both of the coarse-grade projects, AL-3 and AL-5 had lower laboratory air voids at the agency specified Ndesign level. Both projects averaged 3.0 percent air voids. Project GA-1 also had low air voids at the agency specified Ndesign gyrations (1.9 percent). The field notes taken at the time of construction only indicate tender mix problems for one project, NE-4. NE-4 does fail the Ninital requirements. However, construction issues were not commented on at all for many of the projects, so it is possible that there were tender mix problems on other projects. Historically, contractors have found ways to deal with tender mixes in the field. When the Superpave system was first introduced, the Ninitial requirements worked in conjunction with the restricted zone requirements and the fine aggregate angularity requirements to limit the amount of natural sand, or rounded fine aggregate particles in HMA. The restricted zone requirement has been eliminated since it was 168 TABLE 4.17 Summary of Densities, %Gmm, at Ninitial Pine Troxler Project 20-Year ESALs Gradation Ninitial 1 2 3 1 2 3 KY-1 53,706 C 6.0 85.3 84.9 - 85.1 84.5 - KY-3 84,028 F 7.0 88.8 89.1 88.9 88.8 88.6 88.6 AL-6 143,958 F 8.0 90.8 91.0 - 90.6 91.0 - NE-3 365,719 F 7.0 90.5 91.7 90.8 90.5 92.1 91.3 NE-1 383,385 F 7.0 90.4 91.9 - 90.7 91.9 - CO-3 523,624 C 8.0 87.8 88.3 - 88.1 88.3 - CO-4 720,911 F 7.0 88.4 88.9 87.3 88.3 88.4 87.6 CO-1 756,789 F 7.0 90.6 92.3 91.5 90.4 91.9 91.2 UT-1 771,982 F 7.0 87.9 88.8 88.9 87.7 88.5 88.8 FL-1 811,658 C 7.0 85.8 87.9 - 86.1 87.1 - CO-2 1,017,593 F 7.0 91.9 90.8 90.9 91.8 91.0 90.8 CO-5 1,017,593 F 7.0 87.5 87.9 87.6 87.3 87.5 87.4 MI-2 1,250,146 F 7.0 87.8 88.4 87.9 87.9 88.1 87.9 NE-2 1,450,960 F 8.0 89.4 89.6 89.7 89.6 90.1 89.8 MI-3 1,515,200 F 7.0 88.8 89.0 - 88.5 88.8 - AL-5 1,809,675 C 7.0 91.2 91.1 90.9 90.9 90.4 91.0 IN-1 1,850,992 C 8.0 84.3 85.8 85.8 84.3 85.2 85.2 TN-1 3,490,393 F 8.0 89.9 90.2 90.0 91.3 90.8 90.4 AL-2 3,610,001 C 8.0 85.5 84.3 83.9 84.9 83.7 83.4 AL-4 4,899,406 C 8.0 88.6 88.9 89.2 88.7 88.7 89.0 AL-1 6,748,142 C 8.0 86.9 85.9 86.0 87.1 86.1 86.2 GA-1 8,803,521 F 8.0 91.1 91.9 91.8 91.6 92.1 91.4 AL-3 8,861,352 C 8.0 88.9 89.1 - 88.8 89.1 - KS-1 10,075,962 F 8.0 86.4 88.1 87.3 86.7 87.9 87.1 KY-2 12,438,605 C 8.0 81.3 84.8 - 80.9 84.4 - MO-2 12,517,675 C 8.0 - 86.2 84.5 85.6 86.5 84.1 WI-1 14,614,748 C 8.0 87.0 87.5 87.6 86.4 87.5 87.6 MI-1 15,966,398 C 9.0 84.3 85.0 84.2 83.7 84.3 84.0 NE-4 20,084,248 F 8.0 90.1 90.6 90.0 89.7 90.6 90.2 IL-1 26,285,917 C 8.0 83.8 84.5 84.2 84.0 83.9 84.0 MO-1 27,546,007 C 9.0 84.6 85.9 86.1 86.0 86.4 85.7 IL-3 44,466,336 C 8.0 83.6 84.0 83.3 84.7 84.7 84.2 IN-2 45,150,555 C 9.0 88.7 88.5 87.1 88.1 88.4 86.9 IL-2 46,344,297 C 8.0 84.5 86.2 86.1 85.3 87.0 86.8 AR-1 48,726,562 C 9.0 85.0 86.8 86.1 85.0 86.5 86.0 MO-3 53,683,941 C 9.0 85.5 86.5 86.4 85.6 86.4 86.4 NC-1 73,918,507 F 9.0 - 89.3 89.2 89.2 87.7 88.8 AR-2 91,370,805 C 9.0 85.7 85.3 - 85.3 85.5 - AR-4 97,890,077 C 9.0 85.5 86.3 86.2 85.7 85.9 86.3 AR-3 170,842,507 C 9.0 87.5 85.5 - 84.7 86.0 - 169 demonstrated that good performing mixes frequently passed through the restricted zone. Ninitial is sensitive to gradation and the presence of rounded fine aggregate particles. 4.2.8 Evaluation of Nmaximum The densities at Nmaximum, corrected to a DIA of 1.16 degrees, are shown in Table 4.18. AASHTO M 323 specifies that the density at Nmaximum be less than 98 percent. At the agency specified Nmaximum, 36 percent of the Pine samples and 40 percent of the Troxler samples failed the Nmaximum density criteria. One or more samples exceeded the maximum density at Nmaximum for 25 of the 40 projects. When NCAT collected the field data, samples were compacted to both 100 and 160 gyrations. Therefore, sample densities for Nmaximum gyrations greater than 160 gyrations are extrapolated. Although there is a very good relationship between sample density and log of gyrations, at high gyration levels (above the mixtures locking point and Ndesign), this relationship tends to breakdown with additional gyrations producing little increase in sample density. The sample densities at Nmaximum are extrapolated above Nmaximum for 10 of the 25 projects which failed the density requirements at Nmaximum. These extrapolations may be erroneous. However, this still leaves 15 of 40 projects which failed Nmaximum. The maximum rutting for a sample that failed density at Nmaximum occurred for project MI-1, sublot 2, with an average rut depth of 7 mm after four years of traffic. Sublot 1 of MI-1 actually had a slightly higher average rut depth (9 mm) but the sample did not fail the Nmaximum density criteria. Further, as evidenced by Table 4.16, all of the mixes have been extremely rut resistant. Based on the data, the Nmaximum criteria should be eliminated. TABLE 4.18 Summary of Densities, % Gmm, at Nmaximum Pine Troxler Project Nmax 170 1 2 3 1 2 3 AL-1 169 98.5 97.5 97.6 98.9 98.0 97.8 AL-2 160 97.7 97.0 97.2 98.3 96.3 96.4 AL-3 160 97.9 97.7 - 98.0 98.7 - AL-4 160 95.5 96.2 97.0 95.3 96.0 97.1 AL-5 115 98.1 98.0 97.8 98.4 97.8 98.2 FL-1 134 95.6 97.0 - 95.0 96.5 - MI-1 205 97.4 98.5 97.4 96.1 97.4 96.3 MI-2 115 97.9 98.4 98.0 97.3 97.9 97.5 WI-1 160 96.2 97.2 97.3 95.9 96.5 97.0 CO-1 104 98.8 99.9 98.8 98.5 100.0 99.0 CO-2 134 99.4 99.5 99.4 99.5 99.4 99.4 CO-3 174 98.8 98.9 - 98.7 99.2 - CO-4 134 98.7 98.7 97.9 98.3 98.3 97.8 CO-5 134 97.5 97.9 97.8 96.8 97.3 97.1 IN-1 160 97.4 98.7 98.6 96.2 97.6 97.4 IN-2 205 98.2 98.8 97.2 97.5 97.9 96.5 KY-1 75 96.7 96.8 - 95.8 95.5 - KY-2 160 94.9 98.4 - 93.5 97.2 - KY-3 115 97.0 97.5 97.8 96.7 96.8 97.2 AL-6 150 97.2 97.8 - 97.4 98.0 - AR-1 205 97.3 99.5 98.4 97.5 99.0 98.9 AR-2 205 97.5 98.4 - 97.8 98.5 - AR-3 205 96.8 97.8 - 97.3 99.2 - AR-4 205 95.8 96.5 96.3 95.8 96.5 96.9 GA-1 160 98.3 99.1 98.3 98.1 100.0 99.1 IL-1 140 96.5 97.4 97.3 96.3 96.7 96.8 IL-2 140 96.7 98.7 98.5 98.1 99.7 99.4 IL-3 165 95.7 96.6 96.3 97.1 97.6 97.7 KS-1 160 96.9 96.2 96.5 96.2 96.8 96.6 MI-3 115 97.0 97.0 - 96.6 96.5 - MO-1 205 99.0 100.0 100.0 100.0 100.0 100.0 MO-2 160 - 98.5 97.6 98.6 99.3 97.8 MO-3 205 99.5 100.0 100.0 99.9 100.0 100.0 NC-1 205 - 96.5 95.9 96.5 96.3 96.9 NE-1 104 96.5 97.8 - 97.0 98.4 - NE-2 152 97.0 97.6 97.4 97.3 97.9 97.7 NE-3 117 96.6 97.6 96.9 96.9 98.2 97.3 NE-4 174 98.5 98.9 98.4 99.0 99.4 98.5 TN-1 160 97.9 97.7 97.7 98.6 98.5 98.4 UT-1 115 97.9 99.0 98.7 97.4 98.7 98.6 171 4.2.9 Summary and Discussion of Test Results The asphalt content of HMA mixture, as-constructed density and ultimate density are all critical to the performance of an HMA pavement. These values are all interrelated since mixes with higher asphalt contents, for a given aggregate structure, are generally easier to compact initially, and will tend to densify more under traffic. The determination of a HMA mixture?s optimum asphalt content has changed significantly since the first asphalt pavements were introduced in the 1870?s. Optimum asphalt contents were initially selected by experience. As the popularity of HMA grew, there were not enough experienced individuals to determine the optimum asphalt content for all of the HMA being placed. In the late 1930?s and 1940?s, asphalt technologists began to develop laboratory compaction methods with the goal of matching the ultimate pavement density. It had been observed that an HMA pavement densified under traffic from its as- constructed density to an ultimate density, typically within 2 to 3 years after construction. Initially, only one laboratory compaction level was used for a given system, but as tire pressures and traffic volumes grew, the concept of a tiered design system, illustrated in Figure 2.7 (25) was developed where laboratory compaction increased for increasing tire pressures or traffic volumes. The concept of a tiered laboratory compaction was to address the tendency for increased tire pressure, or traffic volumes to produce a denser aggregate skeleton. However, if the laboratory compaction effort was too high, it could be difficult for the contractor to achieve the required as-constructed density in the field. A general summary of the historic HMA mix design philosophy would be to put as much asphalt in a mix as possible without compromising rut resistance. Hveem (5) suggested 172 just enough asphalt to allow adequate compaction in the field with the equipment available. Marshall was quoted as emphasizing the importance of designing the densest (i.e., minimum VMA) possible aggregate structure (6). A tiered system was adopted for the Superpave mix design system. In the Superpave mix design system, minimum required aggregate properties, such as angularity, recommendations for high temperature binder grade, volumetric properties, and laboratory compaction effort all change with design traffic levels. Buchanan (71) demonstrated that for a given gradation, VMA was reduced approximately 1 percent when the Ndesign level was increased by 30 gyrations. Thus, a mixture designed for minimum VMA at an Ndesign level of 125 gyrations would be expected to have a measured VMA of approximately 2 percent above the value at 125 gyrations when compacted to 75 gyrations. Thus, higher Ndesign levels tend to force the aggregate gradation away from the maximum density line. If traffic does not densify these mixtures to as dense of an aggregate structure as the SGC, then the mix gradation may be coarser or finer than is needed. Cooley et al. (81) discussed the influence of gradation on pavement permeability. Coarser mixes tend to be more permeable at a given pavement density than finer mixes are. It is also expected that as the Ndesign level is increased, more compaction effort is required to achieve acceptable density in the field, though this has been difficult to quantify. It should be noted that asphalt content is generally considered to be independent of Ndesign (although dependent for a given mix) and instead dependent on the design (minimum) VMA and air void content. However, Watson et al. (74) indicated that the average design VMA for Georgia DOT mixes, using similar aggregates, was higher for Marshall designed mixes than for Superpave mixes, even though the minimum VMA was the same in both cases. If Ndesign levels are too high, the designer is forced to design closer to the minimum VMA requirement and cannot allow a cushion for production 173 variability. The field data from this study indicated that the as-constructed density, based on cores, for 55 percent of the projects tested was less than 92 percent of Gmm. Statistical analyses indicated that the agency specifications or practices significantly affected the as- constructed density. Two of the agencies with the best as-constructed densities, Colorado and Georgia, have specifications which tend to increase the asphalt content of the mixture. Colorado DOT designs with 100 mm diameter SGC molds. Samples compacted in a 100 mm diameter molds tend to result in lower sample densities as compared to samples compacted in 150 mm diameter molds for the same number of gyrations. Georgia DOT will field-adjust a mixture?s asphalt content in order to ensure specified levels of as-constructed density. The field projects reached their ultimate density after two years of traffic. The majority of the densification occurred in the first three months. The month in which the project was constructed significantly affected the amount of densification which occurred. Projects constructed in the month of May tended to densify the most (approximately 4.0 percent). Projects constructed in April or June on average densified approximately 0.5 percent less than those constructed in May. Projects constructed in July or August densified slightly less than the average of all of the projects, approximately 3.0 percent. Projects constructed in September of October densified the least, an average of approximately 2.3 percent. High temperature PG or the number of 174 high temperature PG bumps as compared to the climatic PG significantly affected pavement densification. Mixes containing PG 76-22 or with two high temperature PG bumps densified less than softer binders. The majority of the samples from the field projects did not achieve the laboratory air void content at the agency specified Ndesign level (Figure 4.17). At a laboratory air void content of 4 percent, the average in-place air void content was 5.5 percent after two-years of traffic. This indicates that the laboratory compaction effort is higher than the combined compaction during construction and from traffic. Brown et al. (78) showed that mixtures designed to 100 gyrations at the 2000 NCAT Test Track compacted to their ultimate density when 10 million ESALs were applied in two years. This equates to more than 100 million ESALs for a 20-year design life, indicating the mixes should have been designed at 125 gyrations using the AASHTO R35-04 Ndesign table. Further, the mixes were designed using an SGC with a low (approximately 1.02) DIA, which would provide less laboratory compaction than an SGC set to a DIA of 1.16 ? 0.02 degrees. Three different analyses were used to try and determine where the Ndesign levels should be set. In the first analysis, the numbers of gyrations to match the 2-year (ultimate) in-place densities were related to the accumulated traffic. The two different compactors used in the study produced back-calculated Ndesign values which differed by approximately 20 gyrations. These differences were attributed to differences in the DIA for the two compactors. This indicates the affect of DIA on the density of laboratory compacted samples. AASHTO (4) has adopted a DIA of 1.16 ? 0.02 degrees as an alternate to an external angle of gyration of 1.25 ? 0.02 degrees. The data were adjusted to a DIA of 1.16 degrees and the resulting back-calculated Ndesign values for the two 175 SGCs compared well (Figure 4.16). A relationship was developed between Log of design traffic (ESALs) and the Log of Ndesign. There was a good deal of scatter in the data, but this was expected based on the literature review. The exclusion of projects constructed with PG 76-22 improved the relationship. Using this relationship the Ndesign values for the currently specified traffic levels could be calculated. The best fit (R 2 = 0.57) indicated reduced gyration levels at all traffic levels (Figure 4.24). The high side of the 80 percent prediction interval approximated the currently specified Ndesign levels. The 80 th percentile for the projects within each category were also calculated; these also indicated reduced Ndesign levels though the reduction in the 0.3 to 1 million ESAL category was minimal. The original Ndesign levels were determined using the best fit of the data, without any adjustment for the confidence or prediction interval (64). However, several projects which could not clearly be identified as outliers were excluded from this analysis and it did not address the use of modified binders. The second analysis looked at the predicted gyrations to match the in-place density at each of the sampling periods (3 months, 6 months, 1 year, 2 years and 4 years). The original Ndesign table was determined by a log-log regression analysis between the gyrations to match the as-constructed density and the density after 12 or more years of traffic and accumulated ESALs (Figure 4.26). This second analysis is then closer to what was originally done to determine the Ndesign levels. This second analysis indicated design gyration levels (Table 4.11) close to those currently specified by AASHTO R 35. However, there is a tremendous amount of scatter in the data (R 2 = 0.37 for Pine Compactor and R 2 = 0.34 for Troxler compactor). 176 The third analysis attempted to reduce the scatter in the data and to adjust the data for the effect of as-constructed density. As noted previously, 55 percent of the projects had as-constructed densities less than 92 percent. It was demonstrated that the as- constructed density affected the 2-year or ultimate density. Models were developed to relate the 2-year percent of laboratory density at 100 gyrations to as-constructed density, high PG grade, and accumulated ESALs. It was found that the predicted gyrations to match a given percentage of laboratory density represented a small range with a standard deviation between 3.44 and 8.99 gyrations. A matrix of expected percentages of laboratory density was developed based on high PG grade and traffic (Table 4.14). The as-constructed density was set to 92 percent in all cases. The number of gyrations to match the percentage of laboratory density determined in the matrix was calculated for each of the projects. An equation was then developed to relate the average gyrations determined to match the in-place densities to high PG grade and traffic, assuming an as- constructed density of 92 percent. Table 4.15 summarizes these results which are similar to the results determined using the first analysis (Table 4.10). Rut depth measurements were taken in the field at the two-year and four-year sampling intervals. A maximum average rut depth for a project after four years of traffic was 7.4 mm with an overall average of 2.7 mm. The rut depth measurements alone support lowering the Ndesign levels since even at 95 percent reliability 2 of 40 pavements would be expected to have unacceptable levels of rutting. Similar findings were reported for the 2000 NCAT Test Track. It was also noted that sections constructed with PG 76-22 at the 2000 NCAT Test Track rutted 60 percent less than sections 177 constructed with PG 67-22. Most of the rutting at the 2000 NCAT Test Track was attributed to pavement densification. Combined, these data indicate that the Ndesign levels can be reduced. As noted previously, the predicted Ndesign levels change very rapidly at 20-year design traffic levels less than 3 million ESALs; therefore, caution must be used in this region. Though lower Ndesign values than currently specified are recommended based on the first analysis for the lowest traffic levels (Table 4.10), there is little or no experience with these levels. Further, density and therefore optimum asphalt content can change very rapidly at lower gyration levels. If the levels are low enough, the compacted samples are not stable immediately after compaction. Therefore, it is recommended that 50 gyrations be maintained for the lowest traffic levels. The combined data from the field projects and the 2000 NCAT Test Track indicate that a maximum Ndesign level of 100 gyrations will provide good performance for very high traffic levels. This is a 25 gyration decrease from the currently specified levels. Table 4.19 summarizes the recommended Ndesign levels for all traffic levels. The values in Table 4.19 are based on Equations 13 and 14. The predicted values from Equation 13 were presented in Table 4.15. The values in Table 4.15 were rounded to produce 4 levels. The largest rounding occurred at 30 million ESALs where the predicted value was 88 and 86 based on Equations 13 and 14, respectively. The recommended Ndesign levels from Table 4.15 are slightly more conservative than the Ndesign levels recommended in Table 4.10. The recommended Ndesign values based on Table 4.15 also account for the effect of PG 76-22. Values are presented for two binder grades, PG 64-22 and PG 76-22. 178 TABLE 4.19 Proposed Ndesign Levels for an SGC DIA of 1.16 ? 0.02 Degrees 20-Year Design Traffic, ESALs 2-Year Design Traffic, ESALs Ndesign Unmodified Ndesign PG 76-22 < 300,000 < 30,000 50 NA 300,000 to 3,000,000 30,000 to 230,000 65 50 3,000,000 to 10,000,000 230,000 to 925,000 80 65 10,000,000 to 30,000,000 925,000 to 2,500,000 80 65 > 30,000,000 > 2,500,000 100 80 In addition to the 20-year design traffic, a two-year design traffic level is shown. The two-year ESALs were used to develop most of the relationships in this study. A 20- year design for a surface course is most likely unreasonably long. Further, the specified traffic growth rate has a large effect on the 20-year design traffic. The WesTrack experiment noted that rate of loading was important, especially for temporary pavements designed for short periods (87). The use of lower Ndesign levels will tend to allow mixtures to be designed with gradations closer to the maximum density line and still meet minimum VMA requirements. The use of lower Ndesign levels will tend to increase optimum asphalt contents slightly since contractors will most likely design with a slightly larger cushion above the minimum specified VMA. However, to ensure the optimum asphalt contents increased, the minimum VMA requirements would also need to be increased. An increase in the minimum VMA requirements of 0.5 percent would result in an increase of approximately 0.2 percent in optimum asphalt content. Thus, the adoption of the recommended Ndesign levels in Table 4.19 along with an increase in minimum VMA of 0.5 percent would have a combined effect of allowing somewhat denser gradations and increasing the optimum asphalt content slightly. 179 CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS The three objectives of this research were 1) to evaluate the field densification of pavements designed using the Superpave mix design system, 2) to verify or determine the correct Ndesign levels, and 3) to evaluate the locking point concept. A wide range of climates, design traffic levels, PG Binder grades, lift thickness to NMAS, gradations and aggregate types were included in this study. The general goal of previous studies to determine the appropriate laboratory compaction effort has been to determine the laboratory compaction effort that matches the ultimate density of the pavement after the application of traffic. Previous studies to determine or confirm laboratory compaction efforts have indicated a great deal of variability between field and laboratory compaction; therefore, variability was expected in this study. The variability in this study may have been acerbated by three factors: 1. Field and traffic compaction are generally constant stress while the SGC is a constant strain device, 2. The mixes sampled in this study contained a wide range of binder grades, not typical of previous studies, 3. The mixes in this study were designed under a tiered system of aggregate properties and Ndesign levels. 180 5.1 CONCLUSIONS Based on the results from this research study, the following conclusions can be made. 1. Pavements appear to reach their ultimate density after two years of traffic. The average in-place density for all of the projects was the same at 2- and 4-years (94.6 percent of Gmm). A fair relationship was determined between the as- constructed density and the density after two years of traffic. The majority of pavement densification, approximately 66 percent, occurs during the first three months after construction. Both the high PG binder grade and the high temperature bumps between the climatic and specified PG were found to significantly affect pavement densification, with stiffer binders resulting in less densification. The ultimate in-place densities of the pavements evaluated in this study were approximately 1.5 percent less than the densities of the laboratory compacted samples at the agency specified Ndesign. 2. The number of gyrations to match the ultimate in-place density was calculated for each project in this study. The calculated values for the two compactors used in this study differed by approximately 20 gyrations. This was attributed to differences in their DIA. The predicted gyrations, adjusted to a DIA of 1.16 degrees showed good agreement between the two machines. 3. A relationship was developed between predicted Ndesign and design traffic for the projects which were not constructed using PG 76-22. Although there was a great deal of scatter in the data, this was expected. The predicted gyration levels were generally less than those currently specified. 181 4. A relationship was also developed to relate the 2-year percent of laboratory density at 100 gyrations to as-constructed density, high PG grade, and accumulated ESALs. It was found that the predicted gyrations to match a given percentage of laboratory density represented a small range with a standard deviation between 3.44 and 8.99 gyrations. A matrix of expected percentages of laboratory density was developed based on high PG grade, traffic and an as- constructed density of 92 percent. The numbers of gyrations to match the percentages of laboratory density determined in the matrix were calculated for all of the projects. An equation was then developed to relate the average gyrations determined to match the in-place densities to high PG grade and traffic. The predicted gyrations were very similar to those determined using the first analysis. However, this analysis accounted for the use of PG 76-22. It was found that Ndesign could be reduced by approximately 15 gyrations when PG 76-22 was specified. 5. All of the projects in this study were very rut resistant. The maximum observed rutting for the field projects was 7.4 mm with an average rut depth for all of the projects of 2.7 mm after 4 years of traffic. 6. The requirements for Ninitial were evaluated based on the field project data. AASHTO M 35 specifies a tiered density requirement at Ninitial depending on traffic level. In the 300,000 to 3,000,000 ESAL range, 32 percent of the samples failed Ninitial requirement. In the greater than 3,000,000 million ESAL range, 20 percent of samples failed Ninitial requirement. The majority of the projects which failed Ninitial were fine-graded. All of the projects are performing well in 182 terms of rutting resistance. Only one project failed Ninital and was tender in the field. There is no strong evidence to keep the requirements for Ninitial 7. The requirement for Nmaximum was evaluated based on the field project data. AASHTO M 35 specifies a density requirement of less than 98 percent at Nmaximum to guard against the potential for rutting. Thirty-six percent of the samples tested with the Pine compactor and 40 percent of the samples tested with the Troxler compactor failed the density requirements at Nmaximum. However, the projects have all been extremely rut resistant. Therefore, the density requirement at Nmaximum does not appear to be a good indicator of rutting potential and should be eliminated. 5.2 RECOMMENDATIONS Based on the research conducted in this study, the following recommendations are made: The specification for angle of gyration should be revised to only allow a DIA of 1.16 ? 0.02 degrees. The Ndesign levels shown in Table 5.1 should be adopted for the design of Superpave HMA. Consideration should be given to the use of the 2-year design traffic volume to determine Ndesign as opposed to the 20-year design traffic volume. The criteria for Ninitial and Nmaximum should be eliminated. TABLE 5.1 Recommended Ndesign Levels for an SGC DIA of 1.16 ? 0.02 Degrees 20-Year Design Traffic, ESALs 2-Year Design Traffic, ESALs Ndesign Unmodified Ndesign PG 76-22 < 300,000 < 30,000 50 NA 300,000 to 3,000,000 30,000 to 230,000 65 50 3,000,000 to 10,000,000 230,000 to 925,000 80 65 10,000,000 to 30,000,000 925,000 to 2,500,000 80 65 > 30,000,000 > 2,500,000 100 80 183 CHAPTER 6 REFERENCES 1. Cominsky, R., ?The Superpave Mix Design Manual for New Construction and Overlays,? SHRP-A-407, Strategic Highway Research Program, National Research Council, Washington, DC, 1994. 2. Brown, E. R., and M. S. Buchanan, ?Superpave Gyratory Compaction Guidelines,? Research Results Digest No. 237, National Cooperative Highway Research Program, Transportation Research Board, National Research Council, Washington, D. C., 1999. 3. Anderson, R. M., R. B. McGennis, W. On Tam, and T. W. Kennedy, ?Sensitivity of Mixture Performance Properties to Changes in Laboratory Compaction Using the Superpave Gyratory Compactor,? In Journal of the Association of Asphalt Paving Technologists, Vol. 69, Reno, NV, 2000, Pp 1-33. 4. American Association of State Highway and Transportation Officials, UStandard Specifications for Transportation Materials and Methods of Sampling and TestingU Part 1B: Specifications,25P th P Ed., Washington, DC, 2005. 5. Hveem, F. N., ?Asphalt Pavements from the Ancient East to the Modern West,? Fifth Annual Nevada Street and Highway Conference, 1970. 6. Leahy, R. B., and R. B. McGennis, ?Asphalt Mixes: Materials, Design and Characterization,? In Journal of the Association of Asphalt Paving Technologists, Vol. 68A, Chicago, IL, 1999, Pp 70-127. 7. Holley *.pdf history of Asphalt Pavements from Internet 8. Halstead, W. J., and J. Y. Welborn, ?History of the Development of Asphalt Testing Apparatus and Asphalt Specifications,? In Proceedings Association of Asphalt Paving Technologists Historical Session, Vol. 43A, Williamsburg, VA, 1974, Pp 89-120. 9. Crawford, C. ?The Rocky Road of Mix Design? Hot Mix Asphalt Technology, Winter 1989, National Asphalt Pavement Association, Riverdale, MD. 10. Davis, R. L. ?Large Stone Mixes: A Historical Insight? Information Series 103, National Asphalt Pavement Association, Riverdale, MD, 1988. 184 11. Roberts, F. L. , P. S. Kandhal, E. R. Brown, D-Y, Lee, and T. W. Kennedy, UHot Mix Asphalt Materials, Mixture Design and ConstructionU Second Edition, NAPA Education Foundation, Lanham, MD, 1996. 12. Vallegra, B. A., and W. R. Lovering, ?Evolution of the Hveem Stabilometer Method of Designing Asphalt Paving Mixtures,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 54, San Antonio, 1985, Pp 243-265. 13. Hveem, F. N. ?Use of the Centrifuge Kerosene Equivalent as Applied to Determine the Required Oil Content for Dense Graded Bituminous Mixes?, In Proceedings of The Association of Asphalt Paving Technologists, Vol. 13, Chicago, 1942, Pp 9-41. 14. War Department, Corps of Engineers, U. S. Army, Mississippi River Commission, ?Investigation of the Design and Control of Asphalt Paving Mixtures,? Technical Memorandum No. 3-254, Vol. 1, U.S. Waterways Experiment Station, Vicksburg, MS, May 1948. 15. White, T. D., ?Marshall Procedures for Design and Quality Control of Asphalt Mixtures,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 54, San Antonio, 1985, Pp 265-284. 16. McLeod, N.W. Relationship Between Density, Bitumen Content, and Voids Properties of Compacted Paving Mixtures. Proceedings of the Highway Research Board, Vol. 35, 1956. 17. McLeod, N.W. Selecting the Aggregate Specific Gravity for Bituminous Paving Mixtures. Proceedings of the Highway Research Board, Vol. 36, 1957. 18. The Asphalt Institute. Mix Design Methods for Asphalt Concrete (MS-2), 2nd ed., May 1963. 19. Foster, C. R., ?Development of Marshall Procedures for Designing Asphalt Paving Mixtures,? Information Series 84, National Asphalt Pavement Association, 1982. 20. Kandhal, P. S., and W. S. Koehler, ?Marshall Mix Design Method: Current Practices,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 54, San Antonio, 1985, Pp 284-303. 21. Ortolani, L., and H. A. Sandberg, Jr., ?The Gyratory-Shear Method of Molding Asphaltitc Concrete Test Specimens; Its Development and Correlation with Field Compaction Methods. A Texas Highway Department Standard Procedure,? In 185 Proceedings of the Association of Asphalt Paving Technologists, Vol. 21, Cincinnati, OH, 1951, Pp 280-297. 22. Philippi, O. A., ?Compaction of Bituminous Concrete,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 26, Atlanta, GA, 1957, Pp 191-201. 23. American Society for Testing and Materials, UAnnual Book of StandardsU, Vol. 4.03 Road and Paving Materials; Vehicle Pavement Systems, West Conshohocken, PA, 2004. 24. Texas Department of Transportation, ?Bituminous Test Procedures Manual,? 200- F, Austin, TX, 2005. 25. McRae, J. L., ?Compaction of Bituminous Concrete,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 26, Atlanta, GA, 1957, Pp 206-213. 26. McRae, J. L., and A. R. McDaniel, ?Progress Report on the Corps of Engineers? Kneading Compactor for Bituminous Mixtures,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 27, Montreal, Quebec, Canada, 1958, Pp- 357-382. 27. Ruth, B. E., X. Shen, L-H, Wang, ?Gyratory Evaluation of Aggregate Blends to Determine their Effect on Shear Resistance and Sensitivity to Asphalt Content, ? UEffects of Aggregates and Mineral Fillers on Asphalt Mixture Performance, ASTM STP 1147U, R. C. Meininger, Ed., American Society for Testing and Materials, Philaelphia, 1992, Pp 252-264. 28. Bonnot, J., ?Asphalt Aggregate Mixtures,? In Transportation Research Record 1096, Transportation Research Board, National Research Council, Washington, D. C., 1986, Pp 42-51. 29. Cominsky, R., R. B. Leahy, and E. T. Harrigan, ?Level One Mix Design: Materials Selection, Compaction, and Conditioning,? SHRP-A-408, Strategic Highway Research Program, National Research Council, Washington, DC, 1994. 30. Hughes, C. S., ?Compaction of Asphalt Pavement,? NCHRP Synthesis No. 152, Transportation Research Board, National Research Council, Washington, DC, 1989. 31. Consuegra, A., D. N. Little, H. Von Quintus, and J. Burati., ?Comparative Evaluation of Laboratory Compaction Devices Based on Their Ability to Produce Mixtures with Engineering Properties Similar to Those Produced in the Field,? In 186 Transportation Research Record 1228, Transportation Research Board, National Research Council, Washington, DC, 1989, Pp 80-87. 32. Von Quintus, H. L., J. A. Scherocman, C. S. Hughes, and T. W. Kennedy, ?Asphalt-Aggregate Mixture Analysis System AAMAS,? NCHRP Report No. 338, Transportation Research Board, National Research Council, Washington, DC, 1991. 33. Sousa, J. B., J. Harvey, L. Painter, J. A. Deacon, and C. L. Monismith, ?Evaluation of Laboratory Procedures for Compacting Asphalt-Aggregate Mixtures,? Report No. SHRP-A/UWP-91-523, Strategic Highway Research Program, National Research Council, Washington, D. C., 1991. 34. Button, J. W., D. N. Little, V. Jagadam, and O. J. Pendleton, ?Correlation of Selected Laboratory Compaction Methods with Field Compaction,? In Transportation Research Record No. 1454, Transportation Research Boards, National Research Council, Washington, D. C. 1994, Pp 193-201. 35. Moultier, R., ?Utilization and Possibilities of the Gyratory Shear Compacting Press,? In Liaison Bulletin of the Bridge and Road Laboratories Special Issue: Bitumens and Bituminous Concrete, 1977. 36. American Association of State Highway and Transportation Officials, AASHTO Provisional Standards. Washington, DC, 2001. 37. Yildirim, Y., M. Solaimanian, R. B. McGennis, and T. W. Kennedy. ?Comparative Analysis of Volumetric Properties for Superpave Gyratory Compactors.? In Transportation Research Record 1712, TRB, National Research Council, Washington, DC. 2000. 38. McGennis, R. B., R. M. Anderson, D. Perdomo and P. Turner. ?Issues Pertaining to Use of Superpave Gyratory Compactor.? In Transportation Research Record 1543, TRB, National Research Council, Washington, DC. 1996. Pp. 139-144. 39. McGennis, R. B., D. Perdomo, T. W. Kennedy, and M. Solaimanian. ?Ruggedness Evaluation of AASHTO TP4: The Superpave Gyratory Compactor, Journal of the Association of Asphalt Paving Technologists, Vol. 66, 1997, pp.277-311. 40. American Association of State Highway and Transportation Officials, AASHTO Provisional Standards. Washington, DC, 2002. 41. ?Application Brief Investigating Measurable Gmb Differences Among Gyratory Compacted Specimens A Round Robin Study.? Troxler Electronic Laboratories, Research Triangle Park, NC, 2001. 187 42. Hinrichsen, J., ?Comparison of Four Brands of Superpave Gyratory Compactors.? In Transportation Research Record 1767, TRB, National Research Council, Washington, DC. 2001. Pp. 167-172. 43. ?Operating Instructions for the Dynamic Angle Validation Kit (DAV) for Superpave Gyratory Compactors.? Test Quip Corporation, New Brighton, MN. 2001. 44. Dalton, F. ?Observations of SGC End Plate Deflection Using the FHWA Angle Validation Kit.? Report 2000-02 Revision C. Pine Instrument Company, PA. 2000. 45. Dalton, F. ?Investigation of Pine Superpave Gyratory Compactors Using the FHWA Angle Validation Kit? Report 2000-03 Revision A. Pine Instrument Company, PA. 2000. 46. Harman, T. ?Dynamic Internal Angle (DIA) Measurement in the Troxler Superpave Gyratory Compactor ? Linearity Study.? Turner-Fairbank Highway Research Center, Federal Highway Administration. McLean, VA. 2001. 47. Dalton, F. ?Application of Internal Angle Verification to Obtain Equivalent Results from Various SGC Models? Report 2001-01 Revision A. Pine Instrument Company, PA. 2001. 48. Al-Khaateeb, G., C. Paugh, K. Stuart, T. Harman, and J. D?Angelo. ?Target and Tolerance for the Angle of Gyration Used in the Superpave Gyratory Compactor (SGC).? In Transportation Research Record 1789, TRB, National Research Council, Washington, DC. 2002. Pp. 208-215. 49. Prowell, B. D., E. R. Brown and M. Huner, UEvaluation of the Internal Angle of Gyration of Superpave Gyratory Compactors in AlabamaU, NCAT 03-04. National Center for Asphalt Technology, 2003. 50. War Department, Corps of Engineers, U. S. Army, Mississippi River Commission, ?Investigation of the Design and Control of Asphalt Paving Mixtures,? Technical Memorandum No. 3-254, Vol. 3, U.S. Waterways Experiment Station, Vicksburg, MS, May 1948. 51. Dillard, J. H. ?Comparison of Density of Marshall Specimens and Pavement Cores.? In Proceeding of the Association of Asphalt Paving Technologists, Vol. 24 Louisville, KY, 1955. pp. 178-232. 52. Campen, W. H., J. R. Smith, L. G. Erickson and L. R. Mertz, ?The Effect of Traffic on the Density of Bituminous Paving Mixtures,? In Proceedings of the 188 Association of Asphalt Paving Technologists, Vol. 30, Charleston, SC, 1960, Pp 378-397. 53. Graham, M. D., W. C. Burnett, J. J. Thomas, and W. C. Dixon. ?Pavement Density-What Influences It.? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 34, Philadelphia, 1965. pp. 286-308. 54. Woodward, E. J., Jr., and J. L. Vicelja, ?Aviation Boulevard ? Evaluation of Materials, Equipment and Construction Procedures,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 34, Philadelphia, PA, 1965, Pp 215-233. 55. Bright, R., B. Steed, J. Steele, and A. Justice, ?The Effect of Viscosity of Asphalt on Properties of Bituminous Wearing Surface Mixtures,? In Proceedings of the Association of Asphalt Paving Technologists, Vol. 35, Minneapolis, MN, 1966, Pp 164-206. 56. Serafin, P. J., L. L. Kole and A. P. Chritz. ?Michigan Bituminous Experimental Road: Final Report.? In Proceedings of The Association of Asphalt Paving Technologists, Vol. 36, Denver, CO, 1967. pp. 582-614. 57. Palmer, R. K. and J. J. Thomas. ?Pavement Density-How It Changes.? In Proceedings of The Association of Asphalt Paving Technologists, Vol. 37, Atlanta, 1968. pp. 542-583. 58. Epps, J. A., B. M. Gallaway and W. W. Scott, Jr. ?Long-Term Compaction of Asphalt Concrete Pavements.? In Highway Research Record, No. 313. Highway Research Board, National Research Council, Washington, DC, 1970. pp. 79-91. 59. Kandahl, P. S., and M. E. Wenger. ?Asphalt Properties in Relation to Pavement Performance.? In Transportation Research Record, No. 544, Transportation Research Board, National Research Council, Washington, DC, 1975. pp. 1-13. 60. Brown, E. R., and S. A. Cross, ?Comparison of Laboratory and Field Density of Asphalt Mixtures,? In Transportation Research Record 1300, Transportation Research Board, National research Council, Washington, D.C., 1991, Pp 1-12. 61. Hanson, D. I., R. B. Mallick and E. R. Brown. ?Five-Year Evaluation of HMA Properties at the AAMAS Test Projects.? In Transportation Research Record, No. 1454, Transportation Research Board, National Research Council, Washington, DC, 1994. pp. 134-143. 62. Stroup-Gardiner, M., D. E. Newcomb, R. Olson, and J. Teig. ?Traffic Densification of Asphalt Concrete Pavements.? In Transportation Research 189 Record, No. 1675, Transportation Research Board, National Research Council, Washington, DC, 1997. pp. 1-9. 63. Brown, E. R., and R. B. Mallick, ?An Initial Evaluation for NB design B Superpave Gyratory Compactor,? In Journal of the Association of Asphalt Paving Technologists, Vol. 67, Boston, MA, 1998, Pp 101-124. 64. Blankenship, P. B., ?Gyratory Compaction Characteristics: Relation to Service Densities of Asphalt Mixtures,? Master?s Thesis, University of Kentucky, Lexington, KY, 1993. 65. Sousa, J.B., G. Way, J. T. Harvey, and M. Hines, ?Comparison of Mix Design Concepts,? In Transportation Research Record 1492, Transportation Research Board, National Research Council, Washington, D.C., 1995, Pp 151-160. 66. Harman, T. P., J. D?Angelo, and J. R. Bukowski, ?Evaluation of SUPERPAVE Gyratory Compactor in the Field Management of Asphalt Mixes: Four Simulation Studies,? In Transportation Research Record 1513, Transportation Research Board, National Research Council, Washington, D. C., 1995, Pp 1-8. 67. Gowda, G., K. Hall, R. Elliot, A. Meadors, ?Critical Evaluation of Superpave Volumetric Mix Design Using Arkansas Surface Course Mixes,? In Journal of the Association of Asphalt Paving Technologists, Vol. 66, Salt Lake City, UT, 1997, Pp 250-276. 68. Habib, A., M. Hossain, R. Kaldate, and G. A. Fager, ?Comparison of Superpave and Marshall Mixtures for Low-Volume Roads and Shoulders,? In Transportation Research Record 1609, Transportation Research Board, National Research Council, Washington, D. C., 1998, Pp 44-50. 69. Mallick, R. B., S. Buchanan, E. R. Brown, and M. Huner, ?Evaluation of Superpave Gyratory Compaction of Hot Mix Asphalt,? In Transportation Research Record 1638, Transportation Research Board, National Research Council, Washington, D. C., 1998, Pp 111-119. 70. Forstie, D. A., and D. K. Corum, ?Determination of Key Gyratory Compaction Points for Superpave Mix Design in Arizona,? In Progress of Superpave (Superior Performing Asphalt Pavement): Evaluation and Implementation, ASTM STP 1322, R. N. Jester, Ed., American Society for Testing and Materials, 1997. Pp 201-208. 71. Buchanan, M. S., ?Evaluation of the Superpave Compaction Procedure,? Doctoral Dissertation, Auburn University, Auburn, AL, 1999. 190 72. American Association of State Highway and Transportation Officials, AASHTO Provisional Standards. Washington, DC, 2001. 73. Harmelink, D. and T. Aschenbrener, UIn-Place Voids Monitoring of Hot Mix Asphalt PavementsU Report No. CDOT-DTD-R-2002-11, Colorado Department of Transportation, Denver, CO, 2002. 74. Watson, D. E., J. Moore and E. R. Brown, UVerification of Superpave NUBU design UBU Compaction LevelsU Phase I Final Report, Georgia Department of Transportation, Atlanta, GA, 2005. 75. Pine, W. J., USuperpave Gyratory Compaction and the Ndesign TableU, Internal Report to Illinois Department of Transportation, Illinois, 1997. 76. Vavrik, W. R., and S. H. Carpenter, ?Calculating Air Voids at Specified Numbers of Gyrations in Superpave Gyratory Compactor,? In Transportation Research Record 1630, Transportation Research Board, National Research Council, Washington, DC, 1998, Pp 117-125. 77. Federal Highway Administration. LTPPBIND, Version 2.1. Washington, D. C. 1999. 78. Brown, E. R., B. D. Prowell, L. A. Cooley, Jr., J. Zhang, and R. B. Powell, Evaluation of Rutting Performance on 2000 NCAT Test Track, In Journal of Association of Asphalt Paving Technologists, Vol. 73, Baton Rouge, LA 2004, Pp 287-336. 79. American Association of State Highway and Transportation Officials, UAASHTO Guide for Design of Pavement Structures 1986U. AASHTO, Washington, D.C. 1986. 80. Brown, E. R. Density of Asphalt Concrete ? How Much Is Needed? Transportation Research Record 1282, Transportation Research Board, National Academy of Sciences, Washington, D.C. 1990. pp. 27-32. 81. Cooley, Jr., L.A., E.R. Brown, and S. Maghsoodloo. Development of Critical Field Permeability and Pavement Density Values for Coarse Graded Superpave Pavements. In Transportation Research Record 1761, Transportation Research Board, National Research Council. 2001. pp 41-49. 82. Linden, R. N., J. P. Mahoney, and N. C. Jackson. Effect of Compaction on Asphalt Concrete Performance. In Transportation Research Record 1217, TRB, National Research Council, Washington, D.C. 1989, Pp 20-28. 191 83. Harvey, J. T., T. Hoover, N. F. Coetzee, W. A. Nokes, and F. C. Rust. CALTRANS Accelerated Pavement Test (CAL/APT) Program-Test Results: 1994-1997. Journal of the Association of Asphalt Paving Technologists, Vol. 67, 1998, Pp 644-689. 84. Brown, E. R., and S. A. Cross, ?A National Study of Rutting in Hot Mix Asphalt (HMA) Pavements,? In Journal of the Association of Asphalt Paving Technologists, Vol. 61, Charleston, SC, 1992, Pp 535-583. 85. Montgomery, D. C., UDesign and Analysis of Experiments,U Fifth Ed., John Wiley & Sons, New York, 2001. 86. Moseley, H. L., G. C. Page, J. A. Musselman, G. A. Sholar, and P. B. Upshaw, Evaluation of Dynamic Angle Validator. In Transportation Research Record 1891, TRB, National Research Council, Washington, D.C. 1891, Pp 191-197. 87. Federal Highway Administration. ?Performance of Coarse-Graded Mixes at WesTrack?Premature Rutting,? Final Report. Washington, D.C. 1998. 192 Appendix Field Project Data TABLE A.1 SGC Data for Project AL-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.549 2.473 2.504 87.1 91.7 94.3 95.9 97.0 97.3 98.2 1-2 2.549 2.472 2.502 87.1 91.6 94.3 95.8 97.0 97.2 98.2 1-3 2.549 2.475 2.514 87.5 92.0 94.5 96.0 97.1 97.7 98.6 AVG 87.2 91.7 94.4 95.9 97.0 97.4 98.3 2-1 2.566 2.472 2.506 86.7 91.2 93.8 95.2 96.3 96.8 97.7 2-2 2.566 2.458 2.493 86.1 90.6 93.3 94.7 95.8 96.2 97.2 2-3 2.566 2.453 2.507 85.7 90.3 93.0 94.5 95.6 96.8 97.7 AVG 86.2 90.7 93.4 94.8 95.9 96.6 97.5 3-1 2.548 2.414 2.488 85.4 89.6 92.2 93.6 94.7 96.8 97.6 3-2 2.548 2.468 2.489 87.2 91.8 94.4 95.8 96.9 96.7 97.7 3-3 2.548 2.443 2.490 86.0 90.6 93.2 94.7 95.9 96.8 97.7 AVG 86.2 90.7 93.3 94.7 95.8 96.8 97.7 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.549 2.450 2.489 86.0 90.6 93.4 94.9 96.1 96.7 97.6 1-2 2.549 2.476 2.502 87.3 91.9 94.6 96.0 97.1 97.2 98.2 1-3 2.549 2.462 2.494 86.7 91.3 94.0 95.4 96.6 96.9 97.8 AVG 86.7 91.3 94.0 95.4 96.6 96.9 97.9 2-1 2.566 2.435 2.490 84.9 89.5 92.2 93.7 94.9 96.0 97.0 2-2 2.566 2.468 2.471 86.4 91.0 93.6 95.0 96.2 95.3 96.3 2-3 2.566 2.445 2.521 85.6 90.0 92.5 94.0 95.3 97.4 98.2 AVG 85.6 90.1 92.8 94.2 95.5 96.2 97.2 3-1 2.548 2.414 2.476 85.4 89.7 92.1 93.6 94.7 96.2 97.2 3-2 2.548 2.438 2.467 85.8 90.4 93.0 94.5 95.7 95.8 96.8 3-3 2.548 2.436 2.478 86.0 90.4 92.9 94.4 95.6 96.2 97.3 AVG 85.8 90.1 92.7 94.2 95.3 96.1 97.1 193 TABLE A.2 SGC Data for Project AL-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.466 2.397 2.430 86.0 91.3 94.5 96.1 97.2 97.8 98.5 1-2 2.466 2.390 2.409 85.6 91.0 94.2 95.9 96.9 96.9 97.7 1-3 2.466 2.387 2.408 85.8 91.1 94.1 95.8 96.8 96.8 97.6 AVG 85.8 91.1 94.3 95.9 97.0 97.1 98.0 2-1 2.455 2.363 2.375 84.7 90.2 93.3 95.1 96.3 95.8 96.7 2-2 2.455 2.357 2.398 84.7 90.0 93.2 94.9 96.0 96.7 97.7 2-3 2.455 2.339 2.396 84.2 89.2 92.3 94.1 95.3 96.8 97.6 AVG 84.5 89.8 92.9 94.7 95.8 96.4 97.3 3-1 2.460 2.359 2.405 84.6 89.9 93.1 94.8 95.9 96.9 97.8 3-2 2.460 2.341 2.396 83.7 89.0 92.2 94.0 95.2 96.6 97.4 3-3 2.460 2.352 2.394 84.3 89.5 92.7 94.5 95.6 96.4 97.3 AVG 84.2 89.5 92.6 94.4 95.6 96.6 97.5 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.466 2.386 2.407 85.4 90.6 93.6 95.5 96.8 96.7 97.6 1-2 2.466 2.370 2.411 83.8 90.1 93.2 95.0 96.1 96.9 97.8 1-3 2.466 2.367 2.410 83.9 89.8 92.9 94.8 96.0 96.8 97.7 AVG 84.4 90.1 93.2 95.1 96.3 96.8 97.7 2-1 2.455 2.326 2.342 83.4 88.5 91.7 93.6 94.7 94.4 95.4 2-2 2.455 2.328 2.342 83.8 88.8 91.9 93.6 94.8 94.4 95.4 2-3 2.455 2.303 2.364 82.5 87.7 90.8 92.6 93.8 95.3 96.3 AVG 83.2 88.4 91.5 93.3 94.5 94.7 95.7 3-1 2.460 2.314 2.345 83.0 88.1 91.1 92.9 94.1 94.4 95.3 3-2 2.460 2.315 2.365 82.8 87.9 91.0 92.8 94.1 95.2 96.1 3-3 2.460 2.313 2.365 82.9 88.0 91.1 92.9 94.0 95.2 96.1 AVG 82.9 88.0 91.1 92.9 94.1 94.9 95.9 194 TABLE A.3 SGC Data for Project AL-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.472 2.396 2.428 89.3 93.0 95.1 96.3 96.9 97.6 98.2 1-2 2.472 2.391 2.423 89.2 93.1 95.0 96.1 96.7 97.5 98.0 1-3 2.472 2.395 2.428 88.9 92.9 95.1 96.1 96.9 97.6 98.2 AVG 89.1 93.0 95.1 96.2 96.8 97.6 98.2 2-1 2.487 2.430 2.439 89.4 93.6 95.8 97.0 97.7 97.5 98.1 2-2 2.487 2.429 2.428 89.7 93.7 95.9 97.0 97.7 97.0 97.6 2-3 2.487 2.429 2.448 89.2 93.4 95.7 96.9 97.7 97.8 98.4 AVG 89.4 93.6 95.8 97.0 97.7 97.4 98.0 3-1 3-2 3-3 AVG Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.472 2.380 2.417 88.5 92.3 94.5 95.5 96.3 97.1 97.8 1-2 2.472 2.373 2.406 88.4 91.9 94.1 95.2 96.0 96.7 97.3 1-3 2.472 2.372 2.400 88.0 91.9 94.1 95.2 96.0 96.5 97.1 AVG 88.3 92.0 94.2 95.3 96.1 96.8 97.4 2-1 2.487 2.412 2.448 88.7 92.7 94.9 96.2 97.0 97.8 98.4 2-2 2.487 2.412 2.436 88.5 92.6 95.0 96.1 97.0 97.3 97.9 2-3 2.487 2.415 2.436 88.7 92.7 95.1 96.3 97.1 97.4 97.9 AVG 88.6 92.7 95.0 96.2 97.0 97.5 98.1 3-1 3-2 3-3 AVG 195 TABLE A.4 SGC Data for Project AL-4 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.472 2.396 2.428 89.3 93.0 95.1 96.3 96.9 97.6 98.2 1-2 2.472 2.391 2.423 89.2 93.1 95.0 96.1 96.7 97.5 98.0 1-3 2.472 2.395 2.428 88.9 92.9 95.1 96.1 96.9 97.6 98.2 AVG 89.1 93.0 95.1 96.2 96.8 97.6 98.2 2-1 2.487 2.430 2.439 89.4 93.6 95.8 97.0 97.7 97.5 98.1 2-2 2.487 2.429 2.428 89.7 93.7 95.9 97.0 97.7 97.0 97.6 2-3 2.487 2.429 2.448 89.2 93.4 95.7 96.9 97.7 97.8 98.4 AVG 89.4 93.6 95.8 97.0 97.7 97.4 98.0 3-1 3-2 3-3 AVG Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.472 2.380 2.417 88.5 92.3 94.5 95.5 96.3 97.1 97.8 1-2 2.472 2.373 2.406 88.4 91.9 94.1 95.2 96.0 96.7 97.3 1-3 2.472 2.372 2.400 88.0 91.9 94.1 95.2 96.0 96.5 97.1 AVG 88.3 92.0 94.2 95.3 96.1 96.8 97.4 2-1 2.487 2.412 2.448 88.7 92.7 94.9 96.2 97.0 97.8 98.4 2-2 2.487 2.412 2.436 88.5 92.6 95.0 96.1 97.0 97.3 97.9 2-3 2.487 2.415 2.436 88.7 92.7 95.1 96.3 97.1 97.4 97.9 AVG 88.6 92.7 95.0 96.2 97.0 97.5 98.1 3-1 3-2 3-3 AVG 196 TABLE A.5 SGC Data for Project AL-5 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.487 2.437 2.458 91.9 95.0 96.6 97.5 98.0 98.5 98.8 1-2 2.487 2.442 2.454 92.0 95.2 96.7 97.6 98.2 98.2 98.7 1-3 2.487 2.439 2.458 91.8 95.0 96.7 97.5 98.1 98.5 98.8 AVG 91.9 95.1 96.7 97.5 98.1 98.4 98.8 2-1 2.493 2.445 2.458 91.9 95.0 96.7 97.5 98.1 98.3 98.6 2-2 2.493 2.441 2.458 91.6 94.9 96.6 97.4 97.9 98.2 98.6 2-3 2.493 2.444 2.462 91.8 95.0 96.7 97.5 98.0 98.4 98.8 AVG 91.8 95.0 96.6 97.5 98.0 98.3 98.6 3-1 2.493 2.426 2.456 91.1 94.2 95.9 96.7 97.3 98.2 98.5 3-2 2.493 2.441 2.461 91.8 95.0 96.7 97.5 97.9 98.3 98.7 3-3 2.493 2.438 2.462 91.7 94.9 96.5 97.3 97.8 98.4 98.8 AVG 91.5 94.7 96.3 97.2 97.7 98.3 98.7 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.487 2.418 2.443 90.9 94.0 95.8 96.6 97.2 97.8 98.2 1-2 2.487 2.406 2.438 90.6 93.6 95.4 96.2 96.7 97.6 98.0 1-3 2.487 2.420 2.489 91.0 94.2 95.9 96.7 97.3 99.6 100.1 AVG 90.8 93.9 95.7 96.5 97.1 98.3 98.8 2-1 2.493 2.370 2.446 88.8 92.0 93.7 94.5 95.1 97.6 98.1 2-2 2.493 2.435 2.444 91.1 94.5 96.3 97.1 97.7 97.5 98.0 2-3 2.493 2.421 2.445 90.7 93.9 95.7 96.5 97.1 97.6 98.1 AVG 90.2 93.5 95.2 96.0 96.6 97.6 98.1 3-1 2.493 2.427 2.440 91.0 94.2 95.9 96.8 97.4 97.4 97.9 3-2 2.493 2.426 2.449 90.6 94.1 95.9 96.7 97.3 97.7 98.2 3-3 2.493 2.426 2.446 90.9 94.1 95.9 96.7 97.3 97.8 98.1 AVG 90.8 94.1 95.9 96.7 97.3 97.6 98.1 197 TABLE A.6 SGC Data for Project AL-6 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.548 2.479 2.488 91.2 94.4 96.1 96.9 97.3 97.3 97.6 1-2 2.548 2.478 2.482 91.1 94.3 96.0 96.8 97.3 97.0 97.4 1-3 2.548 2.478 2.489 91.0 94.3 96.0 96.8 97.3 97.4 97.7 AVG 91.1 94.3 96.0 96.8 97.3 97.2 97.6 2-1 2.530 2.475 2.487 91.5 94.7 96.5 97.3 97.8 98.0 98.3 2-2 2.530 2.470 2.482 91.1 94.5 96.3 97.1 97.6 97.8 98.1 2-3 2.530 2.472 2.485 91.2 94.5 96.3 97.2 97.7 97.9 98.2 AVG 91.3 94.6 96.3 97.2 97.7 97.9 98.2 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.548 2.450 2.471 90.0 93.1 94.8 95.6 96.2 96.5 97.0 1-2 2.548 2.456 2.474 90.3 93.3 95.1 95.9 96.4 96.7 97.1 1-3 2.548 2.454 2.465 90.1 93.2 94.9 95.8 96.3 96.3 96.7 AVG 90.1 93.2 94.9 95.8 96.3 96.5 96.9 2-1 2.530 2.450 2.469 90.5 93.7 95.4 96.2 96.8 97.2 97.6 2-2 2.530 2.450 2.467 90.4 93.6 95.4 96.2 96.8 97.1 97.5 2-3 2.530 2.448 2.468 90.5 93.6 95.4 96.2 96.8 97.1 97.5 AVG 90.5 93.6 95.4 96.2 96.8 97.1 97.5 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 198 TABLE A.7 SGC Data for Project AR-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.437 2.325 2.347 85.2 90.0 92.8 94.5 95.4 95.5 96.3 1-2 2.437 2.311 2.361 84.6 89.4 92.3 93.8 94.8 96.2 96.9 1-3 2.437 2.307 2.331 84.6 89.4 92.2 93.7 94.7 94.9 95.7 AVG 84.8 89.6 92.4 94.0 95.0 95.5 96.3 2-1 2.429 2.363 2.378 86.7 91.8 94.7 96.3 97.3 97.2 97.9 2-2 2.429 2.353 2.380 86.4 91.3 94.3 95.9 96.9 97.3 98.0 2-3 2.429 2.361 0.000 86.8 91.8 94.7 96.2 97.2 0.0 0.0 AVG 86.6 91.6 94.6 96.1 97.1 97.3 97.9 3-1 2.436 2.350 2.370 86.0 91.0 93.9 95.5 96.5 96.5 97.3 3-2 2.436 2.351 2.371 86.1 91.1 94.0 95.5 96.5 96.5 97.3 3-3 2.436 2.334 2.370 85.5 90.5 93.3 94.8 95.8 96.5 97.3 AVG 85.9 90.8 93.7 95.3 96.3 96.5 97.3 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.437 2.309 2.317 84.5 89.4 92.1 93.7 94.7 94.2 95.1 1-2 2.437 2.326 2.330 85.1 89.9 92.8 94.4 95.4 94.8 95.6 1-3 2.437 2.263 2.340 82.2 87.3 90.2 91.8 92.9 95.2 96.0 AVG 84.0 88.8 91.7 93.3 94.4 94.7 95.6 2-1 2.429 2.341 2.363 85.6 90.7 93.6 95.3 96.4 96.5 97.3 2-2 2.429 2.314 2.352 84.9 89.7 92.6 94.1 95.3 96.1 96.8 2-3 2.429 2.345 2.338 85.8 91.1 94.0 95.5 96.5 95.5 96.3 AVG 85.5 90.5 93.4 95.0 96.1 96.0 96.8 3-1 2.436 2.325 2.380 84.8 89.8 92.7 94.3 95.4 96.9 97.7 3-2 2.436 2.329 2.340 84.9 90.0 93.0 94.5 95.6 95.2 96.1 3-3 2.436 2.330 2.364 85.1 90.2 93.1 94.6 95.6 96.4 97.0 AVG 84.9 90.0 92.9 94.5 95.6 96.1 96.9 199 TABLE A.8 SGC Data for Project AR-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.464 2.348 2.379 85.4 90.0 92.8 94.3 95.3 95.8 96.6 1-2 2.464 2.342 2.367 85.0 89.8 92.5 94.0 95.0 95.3 96.1 1-3 2.464 2.373 2.375 86.0 90.8 93.8 95.4 96.3 95.6 96.4 AVG 85.4 90.2 93.0 94.6 95.5 95.6 96.3 2-1 2.448 2.344 2.378 84.9 89.9 93.0 94.7 95.8 96.3 97.1 2-2 2.448 2.348 2.383 85.2 90.3 93.2 94.9 95.9 96.6 97.3 2-3 2.448 2.340 2.384 85.0 90.0 93.0 94.6 95.6 96.6 97.4 AVG 85.0 90.1 93.1 94.7 95.8 96.5 97.3 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.464 2.340 2.362 84.5 89.4 92.3 93.9 95.0 95.0 95.9 1-2 2.464 2.340 2.363 84.4 89.3 92.3 93.9 95.0 95.1 95.9 1-3 2.464 2.327 2.356 84.0 88.8 91.7 93.3 94.4 94.8 95.6 AVG 84.3 89.2 92.1 93.7 94.8 95.0 95.8 2-1 2.448 2.328 2.353 84.3 89.4 92.3 94.0 95.1 95.3 96.1 2-2 2.448 2.340 2.360 84.7 89.8 92.8 94.5 95.6 95.5 96.4 2-3 2.448 2.332 2.370 84.3 89.4 92.5 94.2 95.3 96.0 96.8 AVG 84.4 89.5 92.5 94.2 95.3 95.6 96.4 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 200 TABLE A.9 SGC Data for Project AR-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.426 2.322 2.323 92.6 93.6 94.6 95.2 95.7 95.0 95.8 1-2 2.426 2.296 2.343 84.6 89.4 92.1 93.6 94.6 95.9 96.6 1-3 2.426 2.309 2.329 85.1 89.9 92.7 94.2 95.2 95.2 96.0 AVG 87.4 91.0 93.1 94.4 95.2 95.3 96.1 2-1 2.436 2.338 2.359 85.5 90.6 93.5 95.0 96.0 96.1 96.8 2-2 2.436 2.313 2.343 84.9 89.7 92.5 94.0 95.0 95.5 96.2 2-3 2.436 2.326 0.000 85.4 90.2 93.0 94.5 95.5 0.0 0.0 AVG 85.2 90.2 93.0 94.5 95.5 95.8 96.5 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.426 2.280 2.312 83.6 88.5 91.3 92.9 94.0 94.5 95.3 1-2 2.426 2.289 2.310 83.7 88.7 91.7 93.3 94.4 94.4 95.2 1-3 2.426 2.279 2.316 83.9 88.7 91.5 92.9 93.9 94.8 95.5 AVG 83.7 88.6 91.5 93.0 94.1 94.6 95.3 2-1 2.436 2.331 2.337 85.2 90.2 93.1 94.7 95.7 #DIV/0! 95.9 2-2 2.436 2.321 2.354 84.8 89.8 92.7 94.2 95.3 #DIV/0! 96.6 2-3 2.436 2.325 0.000 84.8 90.0 92.9 94.4 95.4 #DIV/0! 0.0 AVG 85.0 90.0 92.9 94.4 95.5 #DIV/0! 96.3 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 201 TABLE A.10 SGC Data for Project AR-4 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.409 2.251 2.302 85.4 89.1 91.3 92.6 93.4 95.0 95.6 1-2 2.409 2.243 2.298 85.1 88.9 91.1 92.3 93.1 94.7 95.4 1-3 2.409 2.254 2.293 85.6 89.4 91.6 92.8 93.6 94.6 95.2 AVG 85.4 89.1 91.3 92.6 93.4 94.8 95.4 2-1 2.392 2.253 2.294 85.9 89.8 92.1 93.3 94.2 95.2 95.9 2-2 2.392 2.266 2.296 86.6 90.4 92.7 93.9 94.7 95.3 96.0 2-3 2.392 2.255 2.287 85.9 89.8 92.1 93.4 94.3 94.9 95.6 AVG 86.1 90.0 92.3 93.5 94.4 95.2 95.8 3-1 2.401 2.261 2.295 85.9 89.8 92.1 93.4 94.2 94.9 95.6 3-2 2.401 2.275 2.295 86.4 90.4 92.6 94.0 94.8 94.9 95.6 3-3 2.401 2.263 2.298 85.9 89.9 92.2 93.5 94.3 95.1 95.7 AVG 86.0 90.0 92.3 93.6 94.4 95.0 95.6 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.409 2.276 2.277 86.1 90.2 92.4 93.6 94.5 93.9 94.5 1-2 2.409 2.274 2.285 85.8 89.9 92.3 93.6 94.4 94.3 94.9 1-3 2.409 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 86.0 90.1 92.4 93.6 94.4 94.1 94.7 2-1 2.392 2.272 2.278 86.3 90.4 92.8 94.1 95.0 94.6 95.2 2-2 2.392 2.274 2.283 86.1 90.4 92.8 94.2 95.1 94.8 95.4 2-3 2.392 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 86.2 90.4 92.8 94.1 95.0 94.7 95.3 3-1 2.401 2.283 2.284 86.5 90.5 92.8 94.2 95.1 94.6 95.1 3-2 2.401 2.286 2.320 86.7 90.7 93.0 94.3 95.2 95.9 96.6 3-3 2.401 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 86.6 90.6 92.9 94.2 95.1 95.2 95.9 202 TABLE A.11 SGC Data for Project CO-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.451 2.431 2.451 91.7 95.5 97.6 98.6 99.2 99.6 100.0 1-2 2.451 2.417 2.454 91.2 95.1 97.1 98.1 98.6 99.6 100.1 1-3 2.451 2.433 2.443 91.1 95.2 97.5 98.6 99.3 99.3 99.7 AVG 91.3 95.3 97.4 98.4 99.0 99.5 99.9 2-1 2.436 2.444 2.454 93.4 97.3 99.3 100.0 100.3 100.4 100.7 2-2 2.436 2.435 2.454 92.4 97.0 98.8 99.6 100.0 100.5 100.7 2-3 2.436 2.444 2.451 92.7 96.8 98.8 99.8 100.3 100.2 100.6 AVG 92.8 97.0 99.0 99.8 100.2 100.3 100.7 3-1 2.450 2.429 2.431 92.3 96.2 98.1 98.8 99.1 99.1 99.2 3-2 2.450 2.429 2.437 92.0 96.2 98.0 98.8 99.1 99.3 99.5 3-3 2.450 2.431 2.437 92.3 96.2 98.1 98.9 99.2 99.3 99.5 AVG 92.2 96.2 98.1 98.8 99.2 99.2 99.4 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.451 2.409 2.424 89.2 92.8 94.9 97.7 98.3 98.4 98.9 1-2 2.451 2.394 2.427 88.6 92.2 94.3 97.1 97.7 98.4 99.0 1-3 2.451 2.407 2.436 90.7 94.5 96.6 97.6 98.2 98.8 99.4 AVG 89.5 93.2 95.2 97.4 98.1 98.5 99.1 2-1 2.436 2.421 2.441 91.7 95.4 97.5 98.7 99.4 99.8 100.2 2-2 2.436 2.424 2.464 91.7 95.7 97.8 98.9 99.5 100.7 101.1 2-3 2.436 2.425 2.437 92.0 95.9 98.0 98.9 99.5 99.7 100.0 AVG 91.8 95.7 97.8 98.8 99.5 100.1 100.5 3-1 2.450 2.405 2.426 91.0 94.6 96.7 97.6 98.2 98.8 99.0 3-2 2.450 2.407 2.427 91.1 94.9 96.9 97.7 98.2 98.7 99.1 3-3 2.450 2.416 2.426 91.4 95.1 97.1 98.0 98.6 98.7 99.0 AVG 91.1 94.9 96.9 97.8 98.3 98.7 99.0 203 TABLE A.12 SGC Data for Project CO-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.428 2.425 2.428 92.8 97.0 98.9 99.6 99.9 99.9 100.0 1-2 2.428 2.417 2.423 92.4 96.7 98.5 99.2 99.5 99.6 99.8 1-3 2.428 2.421 2.417 93.0 97.3 99.0 99.5 99.7 99.4 99.5 AVG 92.7 97.0 98.8 99.5 99.7 99.6 99.8 2-1 2.449 2.431 2.445 91.4 95.7 97.8 98.7 99.3 99.6 99.8 2-2 2.449 2.431 2.452 91.6 95.8 97.9 98.7 99.3 99.9 100.1 2-3 2.449 2.433 2.448 91.6 95.7 97.8 98.7 99.3 99.8 100.0 AVG 91.5 95.7 97.8 98.7 99.3 99.7 100.0 3-1 2.449 2.434 2.438 91.7 95.9 98.0 99.0 99.4 99.5 99.6 3-2 2.449 2.419 2.447 91.1 95.3 97.4 98.3 98.8 99.7 99.9 3-3 2.449 2.436 2.446 92.0 96.2 98.3 99.1 99.5 99.8 99.9 AVG 91.6 95.8 97.9 98.8 99.2 99.7 99.8 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.428 2.409 2.419 91.8 95.8 97.8 98.7 99.2 99.5 99.6 1-2 2.428 2.407 2.398 91.6 95.6 97.7 98.5 99.1 98.7 98.8 1-3 2.428 2.411 2.393 91.9 95.9 97.9 98.9 99.3 98.4 98.6 AVG 91.8 95.8 97.8 98.7 99.2 98.8 99.0 2-1 2.449 2.427 2.438 91.1 95.0 97.3 98.4 99.1 99.1 99.6 2-2 2.449 2.421 2.423 90.6 94.8 97.1 98.2 98.9 97.7 98.9 2-3 2.449 2.416 2.437 90.6 94.6 96.9 97.9 98.7 99.2 99.5 AVG 90.8 94.8 97.1 98.1 98.9 98.7 99.3 3-1 2.449 2.410 2.427 90.7 94.7 96.8 97.8 98.4 98.8 99.1 3-2 2.449 2.420 2.426 90.9 94.9 97.0 98.1 98.8 98.6 99.1 3-3 2.449 2.409 2.429 90.7 94.6 96.8 97.8 98.4 98.7 99.2 AVG 90.8 94.7 96.9 97.9 98.5 98.7 99.1 204 TABLE A.13 SGC Data for Project CO-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.427 2.326 2.398 87.2 91.4 93.8 95.1 95.8 98.3 98.8 1-2 2.427 2.369 2.386 88.5 93.1 95.6 96.9 97.6 97.9 98.3 1-3 2.427 2.366 2.392 88.6 93.2 95.7 96.8 97.5 98.0 98.6 AVG 88.1 92.6 95.0 96.3 97.0 98.1 98.6 2-1 2.435 2.372 2.396 88.5 92.9 95.4 96.7 97.4 97.9 98.4 2-2 2.435 2.364 2.397 88.5 92.9 95.2 96.4 97.1 98.0 98.4 2-3 2.435 2.379 2.395 88.6 93.2 95.8 97.0 97.7 97.9 98.4 AVG 88.5 93.0 95.5 96.7 97.4 97.9 98.4 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.427 2.338 2.367 87.6 91.8 94.2 95.5 96.3 96.9 97.5 1-2 2.427 2.335 2.369 87.4 91.6 94.1 95.4 96.2 97.0 97.6 1-3 2.427 2.335 2.373 87.6 91.8 94.2 95.4 96.2 97.2 97.8 AVG 87.5 91.7 94.2 95.4 96.3 97.0 97.6 2-1 2.435 2.362 2.383 88.1 92.4 95.0 96.2 97.0 97.3 97.9 2-2 2.435 2.342 2.389 87.4 91.7 94.1 95.4 96.2 97.5 98.1 2-3 2.435 2.368 2.387 87.9 92.5 95.2 96.4 97.2 97.4 98.0 AVG 87.8 92.2 94.8 96.0 96.8 97.4 98.0 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 205 TABLE A.14 SGC Data for Project CO-4 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.501 2.445 2.485 89.2 93.3 95.7 96.9 97.8 98.7 99.4 1-2 2.501 2.453 2.484 89.1 93.5 96.0 97.2 98.1 98.8 99.3 1-3 2.501 2.440 2.485 89.3 93.4 95.7 96.9 97.6 98.7 99.4 AVG 89.2 93.4 95.8 97.0 97.8 98.8 99.3 2-1 2.497 2.452 2.475 89.6 93.8 96.1 97.4 98.2 98.5 99.1 2-2 2.497 2.453 2.473 89.6 93.9 96.3 97.5 98.2 98.4 99.0 2-3 2.497 2.456 2.469 89.7 94.0 96.4 97.6 98.4 98.3 98.9 AVG 89.6 93.9 96.3 97.5 98.3 98.4 99.0 3-1 2.510 2.448 2.470 88.3 92.7 95.3 96.7 97.5 97.8 98.4 3-2 2.510 2.430 2.467 87.7 92.0 94.5 95.9 96.8 97.7 98.3 3-3 2.510 2.444 2.466 88.2 92.6 95.2 96.5 97.4 97.5 98.2 AVG 88.1 92.4 95.0 96.4 97.2 97.7 98.3 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.501 2.425 2.455 88.4 92.5 94.9 96.1 97.0 97.5 98.2 1-2 2.501 2.424 2.455 88.2 92.2 94.8 96.0 96.9 97.5 98.2 1-3 2.501 2.415 2.447 88.0 92.1 94.5 95.7 96.6 97.1 97.8 AVG 88.2 92.3 94.7 95.9 96.8 97.4 98.1 2-1 2.497 2.415 2.442 88.4 92.2 94.6 95.9 96.7 97.1 97.8 2-2 2.497 2.424 2.455 88.7 92.7 95.1 96.3 97.1 97.6 98.3 2-3 2.497 2.414 2.445 88.0 92.1 94.5 95.8 96.7 97.2 97.9 AVG 88.4 92.3 94.7 96.0 96.8 97.3 98.0 3-1 2.510 2.416 2.453 87.5 91.6 94.1 95.3 96.3 97.1 97.7 3-2 2.510 2.427 2.442 87.7 91.9 94.5 95.8 96.7 96.6 97.3 3-3 2.510 2.420 2.434 87.6 91.7 94.2 95.5 96.4 96.3 97.0 AVG 87.6 91.7 94.3 95.5 96.5 96.7 97.3 206 TABLE A.15 SGC Data for Project CO-5 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.451 2.365 2.404 88.1 92.1 94.5 95.7 96.5 97.5 98.1 1-2 2.451 2.358 2.413 87.9 92.0 94.2 95.5 96.2 97.8 98.4 1-3 2.451 2.380 2.409 88.6 92.8 95.0 96.3 97.1 97.6 98.3 AVG 88.2 92.3 94.6 95.8 96.6 97.6 98.3 2-1 2.462 2.396 2.418 88.6 92.8 95.3 96.5 97.3 97.6 98.2 2-2 2.462 2.397 2.425 88.6 92.9 95.3 96.6 97.4 98.0 98.5 2-3 2.462 2.399 2.423 88.7 93.0 95.4 96.7 97.4 97.8 98.4 AVG 88.6 92.9 95.3 96.6 97.4 97.8 98.4 3-1 2.462 2.401 2.418 88.5 92.9 95.4 96.7 97.5 97.6 98.2 3-2 2.462 2.393 2.417 88.3 92.6 95.1 96.4 97.2 97.5 98.2 3-3 2.462 2.391 2.421 88.2 92.6 95.0 96.3 97.1 97.7 98.3 AVG 88.3 92.7 95.1 96.4 97.3 97.6 98.2 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.451 2.340 2.369 87.6 91.3 93.5 94.7 95.5 96.0 96.7 1-2 2.451 2.332 2.367 87.1 91.0 93.2 94.3 95.1 96.0 96.6 1-3 2.451 2.338 2.369 87.2 91.1 93.3 94.5 95.4 96.0 96.7 AVG 87.3 91.1 93.3 94.5 95.3 96.0 96.6 2-1 2.462 2.352 2.397 87.5 91.3 93.5 94.7 95.5 96.7 97.4 2-2 2.462 2.363 2.387 87.5 91.6 94.0 95.2 96.0 96.4 97.0 2-3 2.462 2.360 2.389 87.5 91.4 93.8 95.0 95.9 96.4 97.0 AVG 87.5 91.4 93.8 95.0 95.8 96.5 97.1 3-1 2.462 2.358 2.384 87.3 91.3 93.7 94.9 95.8 96.2 96.8 3-2 2.462 2.361 2.371 87.2 91.4 93.8 95.1 95.9 95.6 96.3 3-3 2.462 2.361 2.386 87.5 91.5 93.8 95.0 95.9 96.2 96.9 AVG 87.4 91.4 93.8 95.0 95.9 96.0 96.7 207 TABLE A.16 SGC Data for Project FL-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.460 2.359 2.362 88.3 92.1 94.2 95.3 95.9 95.5 96.0 1-2 2.460 2.291 2.354 85.6 89.4 91.4 92.5 93.1 95.1 95.7 1-3 2.460 2.346 2.390 87.6 91.5 93.6 94.7 95.4 96.6 97.2 AVG 87.1 91.0 93.1 94.1 94.8 95.7 96.3 2-1 2.450 2.359 2.382 88.0 92.2 94.4 95.5 96.3 96.6 97.2 2-2 2.450 2.363 2.392 88.1 92.3 94.5 95.7 96.4 97.1 97.6 2-3 2.450 2.362 2.390 88.1 92.2 94.5 95.6 96.4 97.0 97.6 AVG 88.1 92.2 94.5 95.6 96.4 96.9 97.5 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.460 2.290 2.304 85.2 89.0 91.2 92.4 93.1 93.1 93.7 1-2 2.460 2.295 2.322 85.4 89.3 91.5 92.6 93.3 93.8 94.4 1-3 2.460 2.328 2.358 86.6 90.5 92.8 93.8 94.6 95.3 95.9 AVG 85.7 89.6 91.8 92.9 93.7 94.1 94.6 2-1 2.450 2.325 2.357 87.2 91.0 93.2 94.2 94.9 95.5 96.2 2-2 2.450 2.329 2.364 87.3 91.1 93.2 94.3 95.1 95.9 96.5 2-3 2.450 2.343 2.326 87.6 91.6 93.8 94.8 95.6 94.4 94.9 AVG 87.4 91.2 93.4 94.5 95.2 95.3 95.9 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 208 TABLE A.17 SGC Data for Project GA-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.540 2.478 2.501 91.4 94.6 96.3 97.1 97.6 98.1 98.5 1-2 2.540 2.483 2.509 91.6 94.8 96.4 97.3 97.8 98.4 98.8 1-3 2.540 2.482 2.506 91.3 94.6 96.3 97.2 97.7 98.2 98.7 AVG 91.4 94.7 96.4 97.2 97.7 98.3 98.6 2-1 2.520 2.485 2.506 91.7 95.2 97.1 98.0 98.6 99.1 99.4 2-2 2.520 2.496 2.505 92.3 95.8 97.7 98.5 99.0 99.1 99.4 2-3 2.520 2.499 2.505 92.4 96.0 97.8 98.7 99.2 99.0 99.4 AVG 92.1 95.7 97.5 98.4 98.9 99.1 99.4 3-1 2.537 2.498 2.497 91.9 95.4 97.1 98.0 98.5 98.0 98.4 3-2 2.537 2.527 2.500 93.2 96.6 98.3 99.1 99.6 98.1 98.5 3-3 2.537 2.490 2.504 91.2 94.8 96.7 97.6 98.1 98.4 98.7 AVG 92.1 95.6 97.4 98.2 98.7 98.2 98.6 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.540 2.471 2.448 91.1 94.0 95.7 96.6 97.3 96.0 96.4 1-2 2.540 2.489 2.493 91.5 94.8 96.5 97.4 98.0 97.7 98.1 1-3 2.540 2.476 2.495 90.8 94.3 96.0 96.9 97.5 97.8 98.2 AVG 91.1 94.4 96.1 97.0 97.6 97.2 97.6 2-1 2.520 2.482 2.504 91.7 95.2 97.0 97.9 98.5 99.0 99.4 2-2 2.520 2.480 2.517 91.5 95.0 96.8 97.8 98.4 99.4 99.9 2-3 2.520 2.485 2.501 91.6 95.3 97.1 98.0 98.6 98.8 99.2 AVG 91.6 95.2 97.0 97.9 98.5 99.1 99.5 3-1 2.537 2.477 2.479 90.5 94.2 96.0 97.1 97.6 97.3 97.7 3-2 2.537 2.474 2.498 90.9 94.3 96.1 96.9 97.5 98.0 98.5 3-3 2.537 2.484 2.520 91.1 94.6 96.4 97.3 97.9 98.8 99.3 AVG 90.8 94.3 96.2 97.1 97.7 98.1 98.5 209 TABLE A.18 SGC Data for Project IL-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.502 2.383 2.435 84.1 89.2 92.4 94.1 95.2 96.5 97.3 1-2 2.502 2.381 2.424 84.0 89.1 92.3 94.0 95.2 96.0 96.9 1-3 2.502 2.384 2.437 84.1 89.3 92.4 94.1 95.3 96.5 97.4 AVG 84.1 89.2 92.4 94.1 95.2 96.3 97.2 2-1 2.499 2.415 2.439 85.0 90.4 93.7 95.6 96.6 96.7 97.6 2-2 2.499 2.404 2.443 84.7 90.1 93.3 95.1 96.2 96.9 97.8 2-3 2.499 2.403 2.446 84.5 89.9 93.2 95.0 96.2 96.9 97.9 AVG 84.7 90.1 93.4 95.2 96.3 96.9 97.7 3-1 2.491 2.398 2.439 84.5 89.9 93.3 95.1 96.3 97.1 97.9 3-2 2.491 2.402 2.431 84.6 90.1 93.4 95.3 96.4 96.7 97.6 3-3 2.491 2.387 2.440 84.2 89.6 92.9 94.7 95.8 97.0 98.0 AVG 84.4 89.9 93.2 95.0 96.2 96.9 97.8 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.502 2.388 2.428 85.1 90.0 92.8 94.4 95.4 96.3 97.0 1-2 2.502 2.396 2.418 84.2 89.7 92.9 94.6 95.8 95.7 96.6 1-3 2.502 2.375 2.422 83.5 88.9 92.0 93.8 94.9 95.9 96.8 AVG 84.3 89.5 92.6 94.3 95.4 96.0 96.8 2-1 2.499 2.383 2.417 84.1 89.3 93.3 94.2 95.4 95.9 96.7 2-2 2.499 2.384 2.436 84.0 89.4 93.3 94.2 95.4 96.6 97.5 2-3 2.499 2.389 2.423 84.2 89.5 92.7 94.4 95.6 96.1 97.0 AVG 84.1 89.4 93.1 94.3 95.5 96.2 97.1 3-1 2.491 2.379 2.423 84.2 89.4 92.6 94.3 95.5 96.4 97.3 3-2 2.491 2.385 2.424 84.2 89.5 93.5 94.5 95.7 96.5 97.3 3-3 2.491 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 84.2 89.5 93.1 94.4 95.6 96.4 97.3 210 TABLE A.19 SGC Data for Project IL-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.446 2.345 2.370 84.8 89.8 93.1 94.7 95.9 96.0 96.9 1-2 2.446 2.338 2.387 84.7 89.6 92.8 94.4 95.6 96.8 97.6 1-3 2.446 2.343 2.377 84.9 89.9 93.0 94.7 95.8 96.3 97.2 AVG 84.8 89.8 92.9 94.6 95.7 96.4 97.2 2-1 2.428 2.372 2.401 86.4 91.7 94.9 96.6 97.7 98.1 98.9 2-2 2.428 2.366 2.395 86.2 91.3 94.6 96.4 97.4 97.8 98.6 2-3 2.428 2.376 2.385 86.7 91.9 95.1 96.8 97.9 97.5 98.2 AVG 86.4 91.6 94.8 96.6 97.7 97.8 98.6 3-1 2.433 2.370 2.405 86.0 91.2 94.5 96.3 97.4 98.1 98.8 3-2 2.433 2.376 2.409 86.4 91.6 94.8 96.6 97.7 98.2 99.0 3-3 2.433 2.382 2.402 86.7 91.9 95.1 96.8 97.9 97.9 98.7 AVG 86.3 91.6 94.8 96.6 0.0 98.1 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.446 2.363 2.393 85.0 90.4 93.6 95.5 96.6 96.9 97.8 1-2 2.446 2.354 2.389 84.8 90.2 93.3 95.0 96.2 96.8 97.7 1-3 2.446 2.353 2.385 84.7 90.0 93.3 95.0 96.2 96.6 97.5 AVG 84.8 90.2 93.4 95.2 96.3 96.8 97.7 2-1 2.428 2.378 0.000 86.5 91.8 95.0 96.7 97.9 #DIV/0! 0.0 2-2 2.428 2.369 0.000 86.1 91.4 94.5 96.3 97.6 #DIV/0! 0.0 2-3 2.428 2.374 2.391 86.8 91.7 94.9 96.7 97.8 97.6 98.5 AVG 86.5 91.6 94.8 96.6 0.0 #DIV/0! 0.0 3-1 2.433 2.381 2.403 86.3 91.7 94.9 96.7 97.9 97.9 98.8 3-2 2.433 2.378 2.404 86.3 91.6 94.8 96.5 97.7 98.0 98.8 3-3 2.433 2.383 2.403 86.3 91.7 95.0 96.7 97.9 97.9 98.8 AVG 86.3 91.7 94.9 96.6 0.0 97.9 0.0 211 TABLE A.20 SGC Data for Project IL-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.505 2.353 2.396 83.9 88.6 91.3 92.9 93.9 94.9 95.6 1-2 2.505 2.336 2.409 83.5 88.1 90.8 92.3 93.3 95.3 96.2 1-3 2.505 2.361 2.400 84.0 88.8 91.7 93.2 94.3 95.0 95.8 AVG 83.8 88.5 91.3 92.8 93.8 95.1 95.9 2-1 2.493 2.377 2.404 84.4 89.4 92.6 94.2 95.3 95.5 96.4 2-2 2.493 2.367 2.386 84.3 89.1 92.1 93.8 94.9 94.9 95.7 2-3 2.493 2.365 2.396 84.3 89.1 92.2 93.8 94.9 95.3 96.1 AVG 84.3 89.2 92.3 93.9 95.1 95.2 96.1 3-1 2.493 2.365 2.404 83.8 88.8 92.0 93.7 94.9 95.7 96.4 3-2 2.493 2.359 2.393 83.6 88.6 91.7 93.5 94.6 95.1 96.0 3-3 2.493 2.352 2.394 83.5 88.4 91.5 93.2 94.3 95.1 96.0 AVG 83.6 88.6 91.7 93.5 94.6 95.3 96.1 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.505 2.379 2.407 84.3 89.2 92.2 93.8 95.0 95.3 96.1 1-2 2.505 2.367 2.409 84.1 88.9 91.8 93.4 94.5 95.3 96.2 1-3 2.505 2.370 2.403 84.3 89.1 91.9 93.5 94.6 95.0 95.9 AVG 84.2 89.1 92.0 93.6 94.7 95.2 96.1 2-1 2.493 2.381 2.407 84.3 89.5 92.5 94.3 95.5 95.6 96.6 2-2 2.493 2.370 2.412 84.2 89.2 92.3 93.9 95.1 95.8 96.8 2-3 2.493 2.371 2.405 84.1 89.1 92.2 93.9 95.1 95.6 96.5 AVG 84.2 89.3 92.3 94.0 95.2 95.7 96.6 3-1 2.493 2.370 2.411 83.6 88.8 92.0 93.8 95.1 95.7 96.7 3-2 2.493 2.368 2.414 83.5 88.7 91.9 93.8 95.0 95.9 96.8 3-3 2.493 2.383 2.418 84.1 89.4 92.6 94.4 95.6 96.1 97.0 AVG 83.7 89.0 92.2 94.0 95.2 95.9 96.8 212 TABLE A.21 SGC Data for Project IN-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.465 2.358 2.411 84.8 89.8 92.8 94.5 95.7 96.9 97.8 1-2 2.465 2.338 2.406 84.0 89.0 92.1 93.8 94.8 96.8 97.6 1-3 2.465 2.375 2.404 85.0 90.3 93.5 95.2 96.3 96.8 97.5 AVG 84.6 89.7 92.8 94.5 95.6 96.8 97.6 2-1 2.469 2.407 2.443 86.0 91.4 94.6 96.3 97.5 98.1 98.9 2-2 2.469 2.408 2.445 86.1 91.5 94.7 96.4 97.5 98.2 99.0 2-3 2.469 2.407 2.445 86.1 91.5 94.6 96.4 97.5 98.2 99.0 AVG 86.1 91.5 94.6 96.4 97.5 98.2 99.0 3-1 2.471 2.409 2.443 86.2 91.5 94.6 96.4 97.5 98.1 98.9 3-2 2.471 2.405 2.445 85.9 91.3 94.5 96.2 97.3 98.1 98.9 3-3 2.471 2.408 2.446 86.1 91.5 94.7 96.3 97.5 98.1 99.0 AVG 86.1 91.4 94.6 96.3 97.4 98.1 98.9 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.465 2.312 2.361 83.8 88.1 91.1 92.6 93.8 94.9 95.8 1-2 2.465 2.327 2.357 84.0 88.7 91.6 93.2 94.4 94.7 95.6 1-3 2.465 2.321 2.351 83.6 88.4 91.4 93.0 94.2 94.5 95.4 AVG 83.8 88.4 91.4 93.0 94.1 94.7 95.6 2-1 2.469 2.357 2.396 85.0 89.7 92.8 94.4 95.5 96.2 97.0 2-2 2.469 2.352 2.396 84.8 89.5 92.5 94.1 95.3 96.2 97.0 2-3 2.469 2.346 2.394 84.2 89.2 92.4 94.0 95.0 96.0 97.0 AVG 84.7 89.5 92.6 94.2 95.2 96.1 97.0 3-1 2.471 2.349 2.389 84.6 89.4 92.4 93.9 95.1 95.8 96.7 3-2 2.471 2.354 2.397 84.7 89.6 92.6 94.1 95.3 96.1 97.0 3-3 2.471 2.356 2.395 84.8 89.6 92.6 94.2 95.3 96.0 96.9 AVG 84.7 89.5 92.6 94.1 95.2 96.0 96.9 213 TABLE A.22 SGC Data for Project IN-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.684 2.575 2.620 88.3 91.9 94.0 95.2 95.9 97.0 97.6 1-2 2.684 2.594 2.614 88.7 92.6 94.7 95.8 96.6 96.8 97.4 1-3 2.684 2.596 2.618 88.7 92.6 94.8 95.9 96.7 96.9 97.5 AVG 88.6 92.4 94.5 95.7 96.4 96.9 97.5 2-1 2.673 2.564 2.626 88.1 91.8 94.0 95.2 95.9 97.6 98.2 2-2 2.673 2.586 2.628 88.5 92.4 94.8 96.0 96.7 97.7 98.3 2-3 2.673 2.584 2.624 88.4 92.4 94.6 95.9 96.7 97.5 98.2 AVG 88.4 92.2 94.4 95.7 96.4 97.6 98.2 3-1 2.698 2.539 2.606 86.4 90.0 92.2 93.4 94.1 95.9 96.6 3-2 2.698 2.574 2.612 87.2 91.2 93.5 94.6 95.4 96.1 96.8 3-3 2.698 2.577 2.608 87.3 91.2 93.5 94.7 95.5 96.0 96.7 AVG 87.0 90.8 93.0 94.2 95.0 96.0 96.7 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.684 2.548 2.585 87.4 90.9 93.0 94.2 94.9 95.7 96.3 1-2 2.684 2.542 2.578 87.3 90.7 92.8 93.9 94.7 95.4 96.1 1-3 2.684 2.551 2.572 87.1 91.0 93.1 94.3 95.0 95.2 95.8 AVG 87.2 90.9 93.0 94.1 94.9 95.4 96.1 2-1 2.673 2.551 2.583 87.6 91.3 93.5 94.6 95.4 96.0 96.6 2-2 2.673 2.541 2.577 87.4 91.0 93.1 94.3 95.1 95.8 96.4 2-3 2.673 2.552 2.576 87.7 91.3 93.5 94.6 95.5 95.7 96.4 AVG 87.5 91.2 93.4 94.5 95.3 95.8 96.5 3-1 2.698 2.523 2.570 85.9 89.5 91.7 92.7 93.5 94.7 95.3 3-2 2.698 2.520 2.569 85.8 89.4 91.5 92.6 93.4 94.5 95.2 3-3 2.698 2.545 2.558 86.6 90.2 92.4 93.5 94.3 94.1 94.8 AVG 86.1 89.7 91.8 92.9 93.7 94.4 95.1 214 TABLE A.23 SGC Data for Project KS-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.435 2.308 2.344 86.7 90.5 92.8 94.0 94.8 95.7 96.3 1-2 2.435 2.308 2.410 86.7 90.6 92.8 94.0 94.8 98.4 99.0 1-3 2.435 2.305 2.345 86.6 90.5 92.8 93.9 94.7 95.7 96.3 AVG 86.7 90.6 92.8 94.0 94.7 96.6 97.2 2-1 2.421 2.340 2.336 88.5 92.5 94.8 95.9 96.7 95.9 96.5 2-2 2.421 2.335 2.339 88.3 92.3 94.5 95.7 96.4 96.0 96.6 2-3 2.421 2.338 2.365 88.3 92.5 94.7 95.8 96.6 97.2 97.7 AVG 88.4 92.4 94.7 95.8 96.6 96.4 96.9 3-1 2.413 2.315 2.340 87.5 91.7 94.0 95.2 95.9 96.5 97.0 3-2 2.413 2.316 2.337 87.6 91.7 94.0 95.2 96.0 96.3 96.9 3-3 2.413 2.308 2.328 87.6 91.6 93.8 94.9 95.6 95.9 96.5 AVG 87.6 91.7 94.0 95.1 95.9 96.2 96.8 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.435 2.316 2.324 86.7 90.8 93.1 94.3 95.1 94.9 95.4 1-2 2.435 2.292 2.326 86.0 89.9 92.2 93.3 94.1 95.0 95.5 1-3 2.435 2.296 2.335 85.9 90.0 92.2 93.5 94.3 95.3 95.9 AVG 86.2 90.2 92.5 93.7 94.5 95.0 95.6 2-1 2.421 2.323 2.331 87.5 91.7 93.9 95.1 96.0 95.7 96.3 2-2 2.421 2.324 2.328 87.8 91.8 94.1 95.2 96.0 95.6 96.2 2-3 2.421 2.305 2.333 86.8 91.0 93.1 94.4 95.2 95.7 96.4 AVG 87.4 91.5 93.7 94.9 95.7 95.7 96.3 3-1 2.413 2.302 2.315 87.0 91.1 93.4 94.6 95.4 95.4 95.9 3-2 2.413 2.287 2.317 86.4 90.5 92.7 94.0 94.8 95.4 96.0 3-3 2.413 2.291 2.317 86.5 90.5 92.8 94.1 94.9 95.4 96.0 AVG 86.6 90.7 93.0 94.2 95.0 95.4 96.0 215 TABLE A.24 SGC Data for Project KY-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.480 2.420 2.440 86.7 91.8 95.0 96.7 97.6 97.8 98.4 1-2 2.480 2.431 2.451 86.9 92.4 95.5 97.1 98.0 98.3 98.8 1-3 2.480 2.434 2.447 87.1 92.6 95.7 97.3 98.1 98.1 98.7 AVG 86.9 92.3 95.4 97.0 97.9 98.1 98.6 2-1 2.453 2.408 2.438 86.5 92.0 95.3 97.1 98.2 98.7 99.4 2-2 2.453 2.411 2.436 86.6 92.2 95.5 97.2 98.3 98.6 99.3 2-3 2.453 2.410 2.435 86.6 92.2 95.4 97.1 98.2 98.6 99.3 AVG 86.6 92.1 95.4 97.1 98.2 98.6 99.3 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.480 2.386 2.408 85.7 90.5 93.6 95.1 96.2 96.4 97.1 1-2 2.480 2.391 2.415 86.0 90.8 93.8 95.3 96.4 96.6 97.4 1-3 2.480 2.383 2.412 85.5 90.5 93.5 95.1 96.1 96.5 97.3 AVG 85.7 90.6 93.7 95.2 96.2 96.5 97.2 2-1 2.453 2.356 2.393 85.1 90.1 93.3 94.9 96.0 96.6 97.6 2-2 2.453 2.362 2.383 85.4 90.4 93.5 95.1 96.3 96.2 97.1 2-3 2.453 2.356 2.390 85.1 90.1 93.2 94.9 96.0 96.6 97.4 AVG 85.2 90.2 93.3 95.0 96.1 96.5 97.4 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 216 TABLE A.25 SGC Data for Project KY-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.488 2.313 2.366 81.3 86.7 90.0 91.7 93.0 94.1 95.1 1-2 2.488 2.319 2.369 81.6 87.0 90.3 92.1 93.2 94.3 95.2 1-3 2.488 2.329 2.373 81.8 87.2 90.6 92.4 93.6 94.4 95.4 AVG 81.6 87.0 90.3 92.1 93.3 94.3 95.2 2-1 2.470 2.412 2.438 85.1 91.0 94.6 96.5 97.7 98.0 98.7 2-2 2.470 2.409 2.441 85.0 91.0 94.6 96.4 97.5 98.1 98.8 2-3 2.470 2.412 2.438 85.2 91.1 94.7 96.5 97.7 97.9 98.7 AVG 85.1 91.0 94.6 96.5 97.6 98.0 98.7 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.488 2.277 2.313 80.5 85.5 88.7 90.3 91.5 92.0 93.0 1-2 2.488 2.279 2.313 80.4 85.4 88.7 90.4 91.6 92.0 93.0 1-3 2.488 2.278 2.311 80.4 85.4 88.6 90.3 91.6 92.0 92.9 AVG 80.4 85.4 88.7 90.4 91.6 92.0 92.9 2-1 2.470 2.359 2.384 83.9 89.1 92.5 94.3 95.5 95.7 96.5 2-2 2.470 2.362 2.388 84.0 89.3 92.6 94.4 95.6 95.8 96.7 2-3 2.470 2.346 2.390 83.9 88.9 92.2 93.8 95.0 95.8 96.8 AVG 83.9 89.1 92.4 94.2 95.4 95.7 96.7 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 217 TABLE A.26 SGC Data for Project KY-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.484 2.411 2.432 89.6 93.4 95.4 96.4 97.1 97.4 97.9 1-2 2.484 2.414 2.432 89.5 93.3 95.3 96.4 97.2 97.4 97.9 1-3 2.484 2.403 2.435 89.3 93.0 95.0 96.1 96.7 97.5 98.0 AVG 89.5 93.2 95.2 96.3 97.0 97.4 97.9 2-1 2.481 2.420 2.441 89.8 93.6 95.7 96.8 97.5 97.9 98.4 2-2 2.481 2.420 2.439 89.8 93.6 95.8 96.9 97.5 97.7 98.3 2-3 2.481 2.420 2.440 89.9 93.7 95.8 96.9 97.5 97.8 98.3 AVG 89.8 93.6 95.7 96.8 97.5 97.8 98.3 3-1 2.486 2.430 2.455 89.8 93.8 95.9 97.1 97.7 98.2 98.8 3-2 2.486 2.420 2.457 89.5 93.3 95.4 96.6 97.3 98.3 98.8 3-3 2.486 2.433 2.457 89.8 93.8 96.0 97.2 97.9 98.2 98.8 AVG 89.7 93.6 95.8 96.9 97.7 98.3 98.8 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.484 2.386 2.399 88.6 92.2 94.2 95.3 96.1 96.0 96.6 1-2 2.484 2.383 2.399 88.7 92.2 94.2 95.2 95.9 96.1 96.6 1-3 2.484 2.387 2.401 88.8 92.3 94.3 95.3 96.1 96.1 96.7 AVG 88.7 92.2 94.3 95.3 96.0 96.0 96.6 2-1 2.481 2.377 2.407 88.4 91.9 94.0 95.1 95.8 96.4 97.0 2-2 2.481 2.378 2.405 88.9 92.0 94.0 95.1 95.8 96.3 96.9 2-3 2.481 2.380 2.407 88.6 92.1 94.1 95.2 95.9 96.5 97.0 AVG 88.6 92.0 94.1 95.1 95.9 96.4 97.0 3-1 2.486 2.395 2.419 88.6 92.3 94.5 95.6 96.3 96.7 97.3 3-2 2.486 2.382 2.423 88.3 91.9 94.0 95.1 95.8 96.9 97.5 3-3 2.486 2.393 2.423 88.6 92.3 94.4 95.5 96.3 96.9 97.5 AVG 88.5 92.1 94.3 95.4 96.1 96.8 97.4 218 TABLE A.27 SGC Data for Project MI-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.478 2.340 2.387 83.9 88.9 91.9 93.4 94.4 95.5 96.3 1-2 2.478 2.353 2.393 84.0 89.2 92.3 93.9 95.0 95.7 96.6 1-3 2.478 2.357 2.385 84.0 89.4 92.4 94.1 95.1 95.4 96.2 AVG 84.0 89.2 92.2 93.8 94.8 95.6 96.4 2-1 2.472 2.355 2.406 84.4 89.7 92.7 94.3 95.3 96.6 97.3 2-2 2.472 2.367 2.390 84.8 90.0 93.2 94.8 95.8 95.9 96.7 2-3 2.472 2.372 2.445 84.9 90.2 93.3 94.9 96.0 98.2 98.9 AVG 84.7 90.0 93.1 94.7 95.7 96.9 97.6 3-1 2.497 2.367 2.421 83.8 89.0 92.1 93.8 94.8 96.2 97.0 3-2 2.497 2.364 2.404 83.7 88.9 92.0 93.6 94.7 95.4 96.3 3-3 2.497 2.376 2.400 84.2 89.4 92.5 94.1 95.2 95.0 96.1 AVG 83.9 89.1 92.2 93.8 94.9 95.5 96.4 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.478 2.298 2.336 82.7 87.4 90.3 91.7 92.7 93.5 94.3 1-2 2.478 2.298 2.342 82.5 87.3 90.2 91.7 92.7 93.8 94.5 1-3 2.478 2.307 2.339 82.8 87.6 90.6 92.0 93.1 93.6 94.4 AVG 82.7 87.4 90.4 91.8 92.9 93.6 94.4 2-1 2.472 2.307 2.366 82.9 87.8 90.8 92.3 93.3 94.9 95.7 2-2 2.472 2.328 2.370 83.6 88.5 91.5 93.1 94.2 95.0 95.9 2-3 2.472 2.325 2.364 83.5 88.4 91.5 93.0 94.1 94.8 95.6 AVG 83.3 88.3 91.3 92.8 93.9 94.9 95.7 3-1 2.497 2.337 2.351 83.2 88.0 91.0 92.5 93.6 93.3 94.2 3-2 2.497 2.334 2.353 83.0 87.9 90.8 92.4 93.5 93.5 94.2 3-3 2.497 2.324 2.363 82.8 87.6 90.5 92.0 93.1 93.9 94.6 AVG 83.0 87.8 90.8 92.3 93.4 93.6 94.3 219 TABLE A.28 SGC Data for Project MI-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.446 2.387 2.416 88.4 92.8 95.4 96.8 97.6 98.1 98.8 1-2 2.446 2.389 2.422 88.5 93.0 95.5 96.8 97.7 98.4 99.0 1-3 2.446 2.396 2.421 88.7 93.2 95.7 97.0 98.0 98.4 99.0 AVG 88.6 93.0 95.6 96.9 97.7 98.3 98.9 2-1 2.440 2.395 2.424 88.9 93.4 95.9 97.3 98.2 98.8 99.3 2-2 2.440 2.402 2.420 89.3 93.8 96.4 97.7 98.4 98.7 99.2 2-3 2.440 2.401 2.421 89.2 93.7 96.2 97.6 98.4 98.8 99.2 AVG 89.1 93.6 96.2 97.5 98.3 98.8 99.2 3-1 2.458 2.403 2.436 88.6 93.0 95.6 96.9 97.8 98.5 99.1 3-2 2.458 2.407 2.433 88.9 93.2 95.8 97.1 97.9 98.5 99.0 3-3 2.458 2.403 2.430 88.1 93.0 95.6 96.9 97.8 98.3 98.9 AVG 88.5 93.1 95.7 97.0 97.8 98.4 99.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.446 2.372 2.377 88.3 92.4 94.9 96.2 97.0 96.5 97.2 1-2 2.446 2.352 2.388 87.5 91.6 94.1 95.3 96.2 96.9 97.6 1-3 2.446 2.356 2.385 87.7 91.7 94.2 95.4 96.3 96.8 97.5 AVG 87.8 91.9 94.4 95.6 96.5 96.7 97.4 2-1 2.440 2.367 2.398 88.3 92.3 94.8 96.1 97.0 97.6 98.3 2-2 2.440 2.367 2.390 88.0 92.3 94.8 96.1 97.0 97.3 98.0 2-3 2.440 2.365 2.395 88.0 92.2 94.7 96.0 96.9 97.4 98.2 AVG 88.1 92.3 94.8 96.1 97.0 97.4 98.1 3-1 2.458 2.370 2.402 87.8 91.8 94.2 95.5 96.4 97.0 97.7 3-2 2.458 2.377 2.400 87.8 92.0 94.5 95.8 96.7 97.0 97.6 3-3 2.458 2.372 2.399 87.9 91.9 94.4 95.6 96.5 96.9 97.6 AVG 87.9 91.9 94.4 95.6 96.5 97.0 97.7 220 TABLE A.29 SGC Data for Project MI-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.468 2.391 2.412 89.5 93.2 95.1 96.2 96.9 97.2 97.7 1-2 2.468 2.397 2.421 89.7 93.4 95.4 96.4 97.1 97.7 98.1 1-3 2.468 2.390 2.418 89.3 93.0 95.1 96.2 96.8 97.5 98.0 AVG 89.5 93.2 95.2 96.3 96.9 97.4 97.9 2-1 2.466 2.378 2.410 89.4 92.8 94.7 95.8 96.4 97.2 97.7 2-2 2.466 2.390 2.414 89.7 93.3 95.3 96.2 96.9 97.5 97.9 2-3 2.466 2.394 2.416 89.9 93.5 95.4 96.4 97.1 97.5 98.0 AVG 89.6 93.2 95.1 96.1 96.8 97.4 97.9 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.468 2.363 2.387 88.6 92.1 94.0 95.1 95.7 96.1 96.7 1-2 2.468 2.359 2.388 88.4 91.9 93.9 94.9 95.6 96.2 96.8 1-3 2.468 2.360 2.384 88.4 91.8 93.9 94.9 95.6 96.0 96.6 AVG 88.4 91.9 94.0 95.0 95.7 96.1 96.7 2-1 2.466 2.361 2.381 88.8 92.2 94.1 95.0 95.7 96.0 96.6 2-2 2.466 2.357 2.380 88.7 92.0 94.0 94.9 95.6 96.0 96.5 2-3 2.466 2.359 2.383 88.7 92.0 94.0 94.9 95.7 96.1 96.6 AVG 88.7 92.1 94.0 94.9 95.7 96.1 96.6 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 221 TABLE A.30 SGC Data for Project MO-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.474 2.388 2.435 84.4 89.8 93.3 95.2 96.5 97.4 98.4 1-2 2.474 2.347 2.440 83.5 88.6 91.9 93.6 94.9 97.7 98.6 1-3 2.474 2.399 2.398 84.9 90.3 93.7 95.7 97.0 95.9 96.9 AVG 84.3 89.6 92.9 94.8 96.1 97.0 98.0 2-1 2.476 2.422 2.454 85.5 91.1 94.5 96.6 97.8 98.2 99.1 2-2 2.476 2.424 2.452 85.9 91.6 94.8 96.7 97.9 98.0 99.0 2-3 2.476 2.416 2.445 85.4 91.0 94.4 96.3 97.6 97.8 98.7 AVG 85.6 91.2 94.6 96.5 97.8 98.0 99.0 3-1 2.485 2.439 2.450 85.7 91.3 94.9 96.9 98.1 97.6 98.6 3-2 2.485 2.423 2.444 85.3 90.7 94.3 96.2 97.5 97.3 98.4 3-3 2.485 2.421 2.454 86.5 91.2 94.3 96.2 97.4 97.8 98.8 AVG 85.8 91.1 94.5 96.4 97.7 97.6 98.6 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.474 2.410 2.435 85.1 90.7 94.1 96.0 97.4 97.4 98.4 1-2 2.474 2.408 2.432 85.1 90.6 94.0 95.9 97.3 97.3 98.3 1-3 2.474 2.396 2.431 84.7 90.2 93.6 95.5 96.8 97.3 98.3 AVG 85.0 90.5 93.9 95.8 97.2 97.3 98.3 2-1 2.476 2.420 2.442 85.4 91.0 94.5 96.4 97.7 97.7 98.6 2-2 2.476 2.406 2.448 85.2 90.7 94.1 95.9 97.2 97.9 98.9 2-3 2.476 2.411 2.423 85.3 90.9 94.2 96.1 97.4 96.9 97.9 AVG 85.3 90.8 94.3 96.1 97.4 97.5 98.5 3-1 2.485 2.401 2.439 84.6 90.0 93.4 95.3 96.6 97.2 98.1 3-2 2.485 2.400 2.441 84.7 90.1 93.5 95.4 96.6 97.2 98.2 3-3 2.485 2.407 2.433 84.7 90.2 93.7 95.6 96.9 96.9 97.9 AVG 84.7 90.1 93.5 95.4 96.7 97.1 98.1 222 TABLE A.31 SGC Data for Project MO-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.360 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1-2 2.360 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1-3 2.360 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2-1 2.376 2.321 2.348 86.5 91.7 94.9 96.6 97.7 98.1 98.8 2-2 2.376 2.316 2.345 86.3 91.5 94.7 96.4 97.5 98.1 98.7 2-3 2.376 2.313 2.359 86.5 91.6 94.7 96.3 97.3 98.6 99.3 AVG 86.4 91.6 94.8 96.4 97.5 98.3 98.9 3-1 2.360 2.260 2.308 84.4 89.6 92.8 94.5 95.8 96.9 97.8 3-2 2.360 2.270 2.319 85.0 90.1 93.3 95.1 96.2 97.4 98.3 3-3 2.360 2.274 2.308 85.1 90.3 93.5 95.2 96.4 97.0 97.8 AVG 84.8 90.0 93.2 94.9 96.1 97.1 98.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.360 2.288 2.312 85.3 90.7 94.0 95.7 96.9 97.0 98.0 1-2 2.360 2.286 2.314 85.0 90.4 93.8 95.6 96.9 97.2 98.1 1-3 2.360 2.289 2.315 84.8 90.7 94.0 95.8 97.0 97.2 98.1 AVG 85.1 90.6 93.9 95.7 96.9 97.1 98.0 2-1 2.376 2.318 2.348 86.0 91.4 94.7 96.5 97.6 98.0 98.8 2-2 2.376 2.306 2.344 85.7 90.9 94.2 95.9 97.1 97.8 98.7 2-3 2.376 2.324 2.347 86.3 91.6 94.9 96.6 97.8 97.9 98.8 AVG 86.0 91.3 94.6 96.3 97.5 97.9 98.8 3-1 2.360 2.215 2.293 82.6 87.8 90.9 92.7 93.9 96.3 97.2 3-2 2.360 2.237 2.292 83.5 88.7 91.8 93.6 94.8 96.2 97.1 3-3 2.360 2.262 2.294 84.5 89.6 92.8 94.7 95.8 96.3 97.2 AVG 83.5 88.7 91.9 93.6 94.8 96.3 97.2 223 TABLE A.32 SGC Data for Project MO-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.444 2.361 2.401 85.1 90.6 93.8 95.6 96.6 97.4 98.2 1-2 2.444 2.361 2.408 85.0 90.4 93.7 95.5 96.6 97.7 98.5 1-3 2.444 2.372 2.399 85.4 90.8 94.1 95.9 97.1 97.4 98.2 AVG 85.2 90.6 93.9 95.6 96.8 97.5 98.3 2-1 2.434 2.382 2.416 86.3 91.7 95.0 96.7 97.9 98.4 99.3 2-2 2.434 2.380 2.414 86.0 91.5 94.8 96.6 97.8 98.3 99.2 2-3 2.434 2.384 2.412 86.1 91.7 95.0 96.8 97.9 98.2 99.1 AVG 86.1 91.6 95.0 96.7 97.9 98.3 99.2 3-1 2.436 2.377 2.415 86.0 91.4 94.7 96.5 97.6 98.3 99.1 3-2 2.436 2.390 2.415 86.2 91.8 95.1 96.9 98.1 98.3 99.1 3-3 2.436 2.381 2.408 86.0 91.5 94.8 96.7 97.7 98.1 98.9 AVG 86.1 91.6 94.9 96.7 97.8 98.2 99.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.444 2.348 2.398 84.4 89.8 93.1 94.9 96.1 97.1 98.1 1-2 2.444 2.345 2.393 84.7 89.9 93.0 94.7 95.9 97.0 97.9 1-3 2.444 2.353 2.395 84.6 90.1 93.3 95.1 96.3 97.1 98.0 AVG 84.6 89.9 93.1 94.9 96.1 97.1 98.0 2-1 2.434 2.374 2.396 85.6 91.1 94.5 96.3 97.5 97.5 98.4 2-2 2.434 2.363 2.401 85.1 90.6 94.0 95.9 97.1 97.7 98.6 2-3 2.434 2.367 2.395 85.4 90.9 94.2 96.0 97.2 97.4 98.4 AVG 85.4 90.9 94.2 96.1 97.3 97.5 98.5 3-1 2.436 2.368 2.393 85.4 90.9 94.3 96.0 97.2 97.3 98.2 3-2 2.436 2.369 2.398 85.5 91.0 94.2 96.1 97.2 97.5 98.4 3-3 2.436 2.366 2.400 85.2 90.7 94.1 95.8 97.1 97.6 98.5 AVG 85.4 90.9 94.2 96.0 97.2 97.5 98.4 224 TABLE A.33 SGC Data for Project NC-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.640 2.529 2.554 89.3 92.6 94.3 95.3 95.8 96.3 96.7 1-2 2.640 2.525 2.542 89.6 92.7 94.3 95.1 95.6 95.9 96.3 1-3 2.640 2.521 2.556 89.1 92.4 94.1 95.0 95.5 96.4 96.8 AVG 89.3 92.6 94.2 95.1 95.6 96.2 96.6 2-1 2.638 2.511 2.522 89.3 92.4 93.9 94.7 95.2 95.2 95.6 2-2 2.638 2.511 2.536 89.2 92.3 93.9 94.7 95.2 95.8 96.1 2-3 2.638 2.507 2.550 89.0 92.1 93.7 94.5 95.0 96.2 96.7 AVG 89.2 92.3 93.9 94.6 95.1 95.7 96.1 3-1 2.649 2.526 2.529 89.4 92.5 94.0 94.9 95.4 95.1 95.5 3-2 2.649 2.509 2.525 88.7 91.8 93.4 94.2 94.7 95.0 95.3 3-3 2.649 2.514 2.515 89.3 92.1 93.7 94.4 94.9 94.5 94.9 AVG 89.1 92.1 93.7 94.5 95.0 94.9 95.2 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.640 2.489 2.495 88.6 91.5 93.0 93.8 94.3 94.2 94.5 1-2 2.640 2.482 2.522 88.4 91.2 92.8 93.5 94.0 95.1 95.5 1-3 2.640 2.505 2.520 88.0 91.5 93.3 94.3 94.9 95.0 95.5 AVG 88.3 91.4 93.0 93.9 94.4 94.8 95.2 2-1 2.638 2.361 2.531 83.9 86.8 88.3 89.0 89.5 95.5 95.9 2-2 2.638 2.492 2.511 88.4 91.5 93.1 94.0 94.5 94.8 95.2 2-3 2.638 2.492 2.523 88.3 91.4 93.1 94.0 94.5 95.3 95.6 AVG 86.8 89.9 91.5 92.3 92.8 95.2 95.6 3-1 2.649 2.512 2.530 88.6 91.8 93.4 94.2 94.8 95.1 95.5 3-2 2.649 2.491 2.538 87.9 91.0 92.6 93.5 94.0 95.3 95.8 3-3 2.649 2.482 2.525 87.5 90.7 92.3 93.1 93.7 94.9 95.3 AVG 88.0 91.2 92.8 93.6 94.2 95.1 95.5 225 TABLE A.34 SGC Data for Project NE-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.414 2.330 2.357 90.8 93.7 95.2 96.0 96.5 97.3 97.6 1-2 2.414 2.336 2.349 91.1 93.9 95.5 96.3 96.8 96.9 97.3 1-3 2.414 2.334 2.354 91.0 93.9 95.4 96.3 96.7 97.1 97.5 AVG 91.0 93.8 95.4 96.2 96.7 97.1 97.5 2-1 2.405 2.356 2.366 92.5 95.4 96.8 97.5 98.0 97.9 98.4 2-2 2.405 2.360 2.372 92.5 95.4 96.9 97.6 98.1 98.3 98.6 2-3 2.405 2.356 2.367 92.4 95.3 96.8 97.5 98.0 98.1 98.4 AVG 92.5 95.4 96.8 97.6 98.0 98.1 98.5 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.414 2.327 2.348 90.6 93.5 95.1 95.9 96.4 96.8 97.3 1-2 2.414 2.329 2.340 90.6 93.5 95.2 96.0 96.5 96.5 96.9 1-3 2.414 2.327 2.342 90.6 93.5 95.0 95.8 96.4 96.6 97.0 AVG 90.6 93.5 95.1 95.9 96.4 96.7 97.1 2-1 2.405 2.352 2.364 91.8 94.9 96.5 97.3 97.8 98.0 98.3 2-2 2.405 2.469 2.361 96.7 99.7 101.3 102.1 102.7 97.8 98.2 2-3 2.405 2.216 2.364 86.7 89.5 90.9 91.7 92.1 98.0 98.3 AVG 91.7 94.7 96.2 97.0 97.5 97.9 98.3 3-1 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-2 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3-3 0.000 0.000 0.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 AVG 0.0 0.0 0.0 0.0 0.0 0.0 0.0 226 TABLE A.35 SGC Data for Project NE-2 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.437 2.347 2.374 89.4 92.8 94.6 95.7 96.3 96.9 97.4 1-2 2.437 2.357 2.377 89.8 93.2 95.0 96.1 96.7 97.0 97.5 1-3 2.437 2.358 2.367 89.9 93.3 95.1 96.1 96.8 96.6 97.1 AVG 89.7 93.1 94.9 95.9 96.6 96.9 97.4 2-1 2.437 2.322 2.386 88.4 91.8 93.7 94.6 95.3 97.4 97.9 2-2 2.437 2.373 2.392 90.6 93.9 95.8 96.8 97.4 97.6 98.2 2-3 2.437 2.369 0.000 90.5 93.8 95.6 96.6 97.2 0.0 0.0 AVG 89.8 93.2 95.0 96.0 96.6 97.5 98.0 3-1 2.443 2.367 2.388 90.0 93.4 95.2 96.2 96.9 97.2 97.7 3-2 2.443 2.365 2.388 89.8 93.2 95.1 96.1 96.8 97.2 97.7 3-3 2.443 2.371 2.391 90.1 93.5 95.4 96.4 97.1 97.4 97.9 AVG 89.9 93.4 95.3 96.2 96.9 97.3 97.8 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.437 2.337 2.361 89.0 92.4 94.2 95.2 95.9 96.4 96.9 1-2 2.437 2.343 2.365 89.2 92.6 94.4 95.5 96.1 96.5 97.0 1-3 2.437 2.340 2.357 89.0 92.4 94.3 95.3 96.0 96.1 96.7 AVG 89.1 92.5 94.3 95.3 96.0 96.3 96.9 2-1 2.437 2.356 2.376 89.6 93.0 94.9 96.0 96.7 97.0 97.5 2-2 2.437 2.358 2.375 89.7 93.1 95.0 96.1 96.8 96.9 97.5 2-3 2.437 2.354 0.000 89.5 93.0 94.9 95.8 96.6 0.0 0.0 AVG 89.6 93.0 94.9 96.0 96.7 96.9 97.5 3-1 2.443 2.355 2.378 89.4 92.8 94.7 95.7 96.4 96.8 97.3 3-2 2.443 2.350 2.374 89.2 92.7 94.5 95.5 96.2 96.7 97.2 3-3 2.443 2.351 2.375 89.1 92.5 94.4 95.5 96.2 96.7 97.2 AVG 89.2 92.7 94.5 95.6 96.3 96.7 97.2 227 TABLE A.36 SGC Data for Project NE-3 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.405 2.317 2.346 91.0 93.7 95.1 95.8 96.3 97.1 97.5 1-2 2.405 2.316 2.341 90.9 93.6 95.0 95.8 96.3 96.9 97.3 1-3 2.405 2.329 2.339 91.5 94.2 95.6 96.4 96.8 96.8 97.3 AVG 91.1 93.8 95.2 96.0 96.5 97.0 97.4 2-1 2.390 2.337 2.350 92.4 95.2 96.6 97.4 97.8 98.0 98.3 2-2 2.390 2.338 2.350 92.6 95.3 96.7 97.4 97.8 97.9 98.3 2-3 2.390 2.321 2.349 91.8 94.5 95.9 96.6 97.1 97.9 98.3 AVG 92.3 95.0 96.4 97.1 97.6 97.9 98.3 3-1 2.398 2.320 2.358 91.6 94.3 95.6 96.3 96.7 98.0 98.3 3-2 2.398 2.322 2.341 91.3 94.3 95.7 96.4 96.8 97.3 97.6 3-3 2.398 2.304 2.341 91.0 93.5 94.9 95.7 96.1 97.3 97.6 AVG 91.3 94.0 95.4 96.1 96.6 97.5 97.9 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.405 2.318 2.341 90.8 93.6 95.1 95.9 96.4 97.0 97.3 1-2 2.405 2.276 2.334 89.3 92.0 93.4 94.1 94.6 96.6 97.0 1-3 2.405 2.315 2.323 90.6 93.5 95.0 95.7 96.3 96.2 96.6 AVG 90.2 93.0 94.5 95.2 95.8 96.6 97.0 2-1 2.390 2.328 2.343 92.0 94.7 96.2 97.0 97.4 97.7 98.0 2-2 2.390 2.326 2.334 91.9 94.7 96.1 96.8 97.3 97.3 97.7 2-3 2.390 2.323 2.347 91.9 94.6 96.0 96.8 97.2 97.9 98.2 AVG 91.9 94.7 96.1 96.9 97.3 97.6 98.0 3-1 2.398 2.316 2.331 91.2 93.9 95.4 96.1 96.6 96.9 97.2 3-2 2.398 2.312 2.325 91.2 93.8 95.2 96.0 96.4 96.5 97.0 3-3 2.398 2.310 2.331 91.1 93.8 95.2 95.8 96.3 96.9 97.2 AVG 91.2 93.8 95.3 96.0 96.4 96.8 97.1 228 TABLE A.37 SGC Data for Project NE-4 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.444 2.386 2.409 90.5 94.0 95.9 97.0 97.6 98.1 98.6 1-2 2.444 2.384 2.408 90.4 93.9 95.9 96.9 97.5 98.0 98.5 1-3 2.444 2.383 2.414 90.3 93.9 95.8 96.8 97.5 98.3 98.8 AVG 90.4 93.9 95.9 96.9 97.6 98.1 98.6 2-1 2.438 2.396 2.407 91.2 94.7 97.0 97.6 98.3 98.3 98.7 2-2 2.438 2.386 2.421 90.8 94.3 96.6 97.2 97.9 98.8 99.3 2-3 2.438 2.385 2.407 90.6 94.2 96.6 97.2 97.8 98.3 98.7 AVG 90.9 94.4 96.7 97.3 98.0 98.5 98.9 3-1 2.449 2.383 2.416 90.1 93.7 95.6 96.6 97.3 98.1 98.7 3-2 2.449 2.394 2.411 90.5 94.1 96.0 97.1 97.8 97.9 98.4 3-3 2.449 2.388 2.415 90.2 93.8 95.8 96.8 97.5 98.1 98.6 AVG 90.2 93.8 95.8 96.8 97.5 98.1 98.6 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.444 2.357 2.395 89.1 92.7 94.6 95.7 96.4 97.5 98.0 1-2 2.444 2.354 2.403 89.1 92.6 94.5 95.6 96.3 97.8 98.3 1-3 2.444 2.366 2.405 89.4 92.9 94.9 96.1 96.8 97.9 98.4 AVG 89.2 92.7 94.7 95.8 96.5 97.7 98.2 2-1 2.438 2.374 2.402 90.1 93.6 95.6 96.7 97.4 98.0 98.5 2-2 2.438 2.368 2.410 89.9 93.4 95.3 96.4 97.1 98.3 98.9 2-3 2.438 2.383 2.406 90.3 93.9 96.0 97.0 97.7 98.2 98.7 AVG 90.1 93.7 95.6 96.7 97.4 98.2 98.7 3-1 2.449 2.378 2.404 89.6 93.3 95.3 96.3 97.1 97.6 98.2 3-2 2.449 2.379 2.386 89.7 93.3 95.4 96.5 97.1 96.8 97.4 3-3 2.449 2.382 2.393 89.8 93.4 95.5 96.5 97.3 97.2 97.7 AVG 89.7 93.3 95.4 96.4 97.2 97.2 97.8 229 TABLE A.38 SGC Data for Project TN-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.459 2.388 2.415 90.0 93.6 95.5 96.5 97.1 97.7 98.2 1-2 2.459 2.392 2.413 90.4 93.8 95.7 96.7 97.3 97.6 98.1 1-3 2.459 2.389 2.418 90.2 93.7 95.6 96.6 97.2 97.8 98.3 AVG 90.2 93.7 95.6 96.6 97.2 97.7 98.2 2-1 2.467 2.403 2.420 90.3 93.9 95.8 96.8 97.4 97.6 98.1 2-2 2.467 2.404 2.416 90.6 94.0 95.9 96.8 97.4 97.4 97.9 2-3 2.467 2.400 2.419 90.6 93.9 95.8 96.7 97.3 97.5 98.1 AVG 90.5 93.9 95.8 96.8 97.4 97.5 98.0 3-1 2.464 2.398 2.412 90.3 93.8 95.6 96.6 97.3 97.5 97.9 3-2 2.464 2.397 2.413 90.3 93.8 95.6 96.6 97.3 97.4 97.9 3-3 2.464 2.398 2.420 90.2 93.8 95.7 96.7 97.3 97.7 98.2 AVG 90.3 93.8 95.6 96.7 97.3 97.5 98.0 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.459 2.384 2.415 90.8 93.9 95.6 96.4 96.9 97.7 98.2 1-2 2.459 2.392 2.406 90.3 93.7 95.6 96.6 97.3 97.4 97.8 1-3 2.459 2.394 2.411 91.2 94.3 96.0 96.8 97.4 97.6 98.0 AVG 90.8 94.0 95.7 96.6 97.2 97.6 98.0 2-1 2.467 2.399 2.418 90.1 93.7 95.6 96.6 97.2 97.5 98.0 2-2 2.467 2.398 2.416 90.3 93.8 95.5 96.5 97.2 97.4 97.9 2-3 2.467 2.399 2.419 90.4 93.8 95.7 96.3 97.2 97.5 98.1 AVG 90.3 93.7 95.6 96.5 97.2 97.5 98.0 3-1 2.464 2.396 2.417 90.1 93.6 95.5 96.5 97.2 97.6 98.1 3-2 2.464 2.381 2.409 89.5 93.0 94.9 96.0 96.6 97.3 97.8 3-3 2.464 2.394 2.408 90.0 93.6 95.5 96.4 97.2 97.2 97.7 AVG 89.8 93.4 95.3 96.3 97.0 97.4 97.9 230 TABLE A.39 SGC Data for Project UT-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.470 2.410 2.441 88.4 92.9 95.4 96.7 97.6 98.3 98.8 1-2 2.470 2.418 2.442 88.8 93.3 95.8 97.1 97.9 98.3 98.9 1-3 2.470 2.413 2.441 88.7 93.1 95.7 96.9 97.7 98.4 98.8 AVG 88.6 93.1 95.6 96.9 97.7 98.3 98.8 2-1 2.458 2.428 2.445 89.5 94.2 96.8 97.2 98.8 99.0 99.5 2-2 2.458 2.427 2.445 89.7 94.4 96.9 98.1 98.7 99.0 99.5 2-3 2.458 2.432 2.446 89.8 94.5 97.1 98.3 98.9 99.1 99.5 AVG 89.7 94.3 96.9 97.8 98.8 99.0 99.5 3-1 2.465 2.436 2.451 89.7 94.3 96.9 98.1 98.8 99.0 99.4 3-2 2.465 2.432 2.449 89.9 94.4 96.9 98.1 98.7 98.9 99.4 3-3 2.465 2.430 2.449 89.5 94.2 96.8 98.0 98.6 99.0 99.4 AVG 89.7 94.3 96.9 98.1 98.7 99.0 99.4 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.470 2.388 2.410 87.9 92.1 94.6 95.9 96.7 97.0 97.6 1-2 2.470 2.383 2.412 87.6 91.8 94.3 95.6 96.5 97.1 97.7 1-3 2.470 2.374 2.415 87.6 91.7 94.1 95.3 96.1 97.2 97.8 AVG 87.7 91.9 94.3 95.6 96.4 97.1 97.7 2-1 2.458 2.391 2.428 88.3 92.6 95.0 96.4 97.3 98.3 98.8 2-2 2.458 2.405 2.423 88.6 93.1 96.3 97.0 97.8 0.0 98.6 2-3 2.458 2.394 2.424 88.4 92.7 95.2 96.6 97.4 98.0 98.6 AVG 88.4 92.8 95.5 96.7 97.5 98.1 98.7 3-1 2.465 2.407 2.433 88.9 93.2 95.6 96.9 97.6 98.1 98.7 3-2 2.465 2.412 2.428 88.7 93.2 95.8 97.1 97.8 97.9 98.5 3-3 2.465 2.404 2.422 88.8 93.0 95.4 96.8 97.5 97.7 98.3 AVG 88.8 93.1 95.6 96.9 97.7 97.9 98.5 231 TABLE A.40 SGC Data for Project WI-1 Sample Gmm Gmb's %Gmm - Gyratory A @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.563 2.454 2.475 87.4 91.5 93.8 95.0 95.7 95.9 96.6 1-2 2.563 2.443 2.490 86.9 91.0 93.3 94.5 95.3 96.5 97.2 1-3 2.563 2.453 2.457 87.5 91.4 93.7 95.0 95.7 95.2 95.9 AVG 87.3 91.3 93.6 94.8 95.6 95.9 96.5 2-1 2.558 2.459 2.490 87.9 91.9 94.2 95.4 96.1 96.7 97.3 2-2 2.558 2.456 2.495 87.7 91.8 94.0 95.3 96.0 96.9 97.5 2-3 2.558 2.458 2.494 87.6 91.8 94.0 95.3 96.1 96.9 97.5 AVG 87.7 91.8 94.1 95.3 96.1 96.9 97.5 3-1 2.546 2.451 2.486 87.5 91.7 94.1 95.4 96.3 97.0 97.6 3-2 2.546 2.466 2.474 88.2 92.5 94.9 96.1 96.9 96.5 97.2 3-3 2.546 2.453 2.490 87.9 92.0 94.3 95.6 96.3 97.2 97.8 AVG 87.9 92.1 94.4 95.7 96.5 96.9 97.5 Sample Gmm Gmb's %Gmm - Gyratory B @ 100 @ 160 @ 8 @ 25 @ 50 @ 75 @ 100 @ 125 @ 160 (Ndesign) (Nmax) (Ninitial) (Ndesign) (Nmax) 1-1 2.563 2.405 2.447 85.7 89.5 91.8 92.9 93.8 94.8 95.5 1-2 2.563 2.411 2.446 86.1 89.9 92.1 93.3 94.1 94.8 95.4 1-3 2.563 2.414 2.435 86.1 89.9 92.2 93.3 94.2 94.3 95.0 AVG 85.9 89.7 92.1 93.2 94.0 94.6 95.3 2-1 2.558 2.433 2.452 87.1 90.9 93.2 94.3 95.1 95.2 95.9 2-2 2.558 2.434 2.459 87.0 90.8 93.1 94.3 95.2 95.5 96.1 2-3 2.558 2.429 2.454 87.0 90.7 92.9 94.1 95.0 95.3 95.9 AVG 87.0 90.8 93.1 94.3 95.1 95.3 96.0 3-1 2.546 2.425 2.460 87.0 90.9 93.2 94.3 95.2 96.0 96.6 3-2 2.546 2.426 2.449 86.9 90.8 93.2 94.4 95.3 95.5 96.2 3-3 2.546 2.435 2.455 87.3 91.3 93.7 94.8 95.6 95.8 96.4 AVG 87.1 91.0 93.4 94.5 95.4 95.8 96.4 232 233 TAB L E A.4 1 C o re Dat a fo r P r o j ect AL - 1 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .549 2.202 2.3 9 1 2 .440 2.39 8 2 .406 0.000 8 6 .4 93.8 9 5 . 7 94.1 9 4 . 4 0 .0 1 - 2 2 .549 2.259 2.3 9 9 2 .396 2.38 4 2 .397 0.000 8 8 .6 94.1 9 4 . 0 93.5 9 4 . 0 0 .0 1 - 3 2 .549 2.281 2.3 9 5 2 .421 2.37 0 2 .398 0.000 8 9 .5 94.0 9 5 . 0 93.0 9 4 . 1 0 .0 Av g. 8 8 .2 94.0 9 4 . 9 93.5 9 4 . 2 0 .0 St d. 1 . 60 0.16 0.8 7 0.55 0. 19 0.00 2 - 1 2 .566 2.333 2.3 9 3 2 .386 2.38 8 2 .420 2.431 9 0 .9 93.3 9 3 . 0 93.1 9 4 . 3 94.7 2 - 2 2 .566 2.283 2.3 4 8 2 .361 2.35 5 2 .368 2.389 8 9 .0 91.5 9 2 . 0 91.8 9 2 . 3 93.1 2 - 3 2 .566 2.278 2.3 5 9 2 .381 2.35 7 2 .392 2.423 8 8 .8 91.9 9 2 . 8 91.9 9 3 . 2 94.4 Av g. 8 9 .6 92.2 9 2 . 6 92.2 9 3 . 3 94.1 St d. 1 . 19 0.91 0.5 2 0.72 1. 01 0.87 3 - 1 2 .548 2.256 2.3 8 6 2 .412 2.40 2 2 .416 2.430 8 8 .5 93.6 9 4 . 7 94.3 9 4 . 8 95.4 3 - 2 2 .548 2.256 2.3 6 1 2 .366 2.35 5 2 .408 2.392 8 8 .5 92.7 9 2 . 9 92.4 9 4 . 5 93.9 3 - 3 2 .548 2.244 2.4 0 1 2 .362 2.36 2 2 .378 2.402 8 8 .1 94.2 9 2 . 7 92.7 9 3 . 3 94.3 Av g. 8 8 .4 93.5 9 3 . 4 93.1 9 4 . 2 94.5 St d. 0 . 27 0.79 1.0 9 1.00 0. 79 0.77 Roadw a y Core - Gm b R oadw ay C o re - % G mm 234 TAB L E A.4 2 C o re Dat a fo r P r o j ect AL - 2 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .466 2.182 2.1 9 6 2 .204 2.19 4 2 .186 2.199 8 8 .5 89.1 8 9 . 4 89.0 8 8 . 6 89.2 1 - 2 2 .466 2.142 2.1 6 4 2 .185 B R OKE N 2 . 383 2.238 8 6 .9 87.8 8 8 . 6 0 .0 96 .6 90.8 1 - 3 2 .466 2.184 2.1 7 9 2 .213 2.18 5 2 .215 2.237 8 8 .6 88.4 8 9 . 7 88.6 8 9 . 8 90.7 Av g. 8 8 .0 88.4 8 9 . 2 88.8 9 1 . 7 90.2 St d. 0 . 96 0.65 0.5 8 0.26 4. 31 0.90 2 - 1 2 .455 2.176 2.2 1 4 2 .206 2.20 0 2 .226 2.256 8 8 .6 90.2 8 9 . 9 89.6 9 0 . 7 91.9 2 - 2 2 .455 2.179 2.2 3 8 2 .210 B R OKE N 2 . 224 2.231 8 8 .8 91.2 9 0 . 0 0 .0 90 .6 90.9 2 - 3 2 .455 2.169 2.2 3 0 2 .201 B R OKE N 2 . 215 2.232 8 8 .4 90.8 8 9 . 7 0 .0 90 .2 90.9 Av g. 8 8 .6 90.7 8 9 . 8 89.6 9 0 . 5 91.2 St d. 0 . 21 0.50 0.1 8 0.00 0. 24 0.58 3 - 1 2 .460 2.155 2.2 2 8 2 .234 2.25 1 2 .276 2.292 8 7 .6 90.6 9 0 . 8 91.5 9 2 . 5 93.2 3 - 2 2 .460 2.179 2.2 5 9 2 .242 2.27 1 2 .311 2.320 8 8 .6 91.8 9 1 . 1 92.3 9 3 . 9 94.3 3 - 3 2 .460 2.183 2.2 9 6 2 .274 2.28 5 2 .283 2.292 8 8 .7 93.3 9 2 . 4 92.9 9 2 . 8 93.2 Av g. 8 8 .3 91.9 9 1 . 5 92.2 9 3 . 1 93.6 St d. 0 . 62 1.38 0.8 6 0.69 0. 75 0.66 Roadw a y Core - Gm b R oadw ay C o re - % G mm 235 TAB L E A.43 C o re Data f o r Project AL -3 Samp le G m m R o adw ay Co re - Gmb R oad w ay Co re - % G mm In-Pl a c e 3-Month 6 -Month 1 -Y ear 2-Y ear In- P l a c e 3-Month 6 -Month 1 - Y ear 2-Y ear 1- 1 2.472 2.190 2.277 2.285 2.243 2.301 88.6 92.1 92.4 90.7 93.1 1- 2 2.472 2.217 2.292 2.315 2.307 2.317 89.7 92.7 93.6 93.3 93.7 1- 3 2.472 2.204 2.285 2.285 2.314 2.307 89.2 92.4 92.4 93.6 93.3 AV G 89.1 92.4 92.8 92.6 93.4 2- 1 2.487 2.259 2.330 2.350 2.349 2.354 90.8 93.7 94.5 94.5 94.7 2- 2 2.487 2.232 2.332 2.319 2.338 2.334 89.7 93.8 93.2 94.0 93.8 2- 3 2.487 2.238 2.290 2.316 2.328 2.315 90.0 92.1 93.1 93.6 93.1 AV G 90.2 93.2 93.6 94.0 93.9 3- 1 3- 2 3- 3 AV G 236 TAB L E A.44 C o re Data f o r Project AL -4 Samp le G m m R o a dw ay Co re - Gmb R oad w ay Co re - % G mm In-Pl a c e 3-Month 6 -Month 1 -Y ear 2-Y ear In- P l a c e 3-Month 6 -Month 1 - Y ear 2-Y ear 1- 1 2 .525 2.238 2.300 2.327 2.339 2.366 88.6 91.1 92.2 92.6 93.7 1- 2 2 .525 2.234 2.319 2.326 2.325 2.353 88.5 91.8 92.1 92.1 93.2 1- 3 2 .525 2.189 2.328 2.339 2.336 2.363 86.7 92.2 92.6 92.5 93.6 AV G 87.9 91.7 92.3 92.4 93.5 2- 1 2 .528 2.199 2.353 2.366 2.348 2.377 87.0 93.1 93.6 92.9 94.0 2- 2 2 .528 2.185 2.341 2.345 2.338 2.423 86.4 92.6 92.8 92.5 95.8 2- 3 2 .528 2.248 2.321 2.330 2.316 2.356 88.9 91.8 92.2 91.6 93.2 AV G 87.4 92.5 92.8 92.3 94.4 3- 1 2 .514 2.238 2.365 2.359 2.348 2.392 89.0 94.1 93.8 93.4 95.1 3- 2 2 .514 2.224 2.360 2.351 2.348 2.380 88.5 93.9 93.5 93.4 94.7 3- 3 2 .514 2.302 2.381 2.382 2.332 2.404 91.6 94.7 94.7 92.8 95.6 AV G 89.7 94.2 94.0 93.2 95.1 237 TAB L E A.45 C o re Data f o r Project AL -5 Samp le G m m R o adw ay Co re - Gmb R oad w ay Co re - % G mm In-Pl a c e 3-Month 6 -Month 1 -Y ear 2-Y ear In- P l a c e 3-Month 6 -Month 1 - Y ear 2-Y ear 1- 1 2.487 2.294 2.362 2.354 2.339 2.363 92.2 95.0 94.7 94.0 95.0 1- 2 2.487 2.281 2.344 2.347 2.345 2.374 91.7 94.3 94.4 94.3 95.5 1- 3 2.487 2.186 2.308 2.324 2.292 2.344 87.9 92.8 93.4 92.2 94.3 AV G 90.6 94.0 94.2 93.5 94.9 2- 1 2.493 2.229 2.318 2.330 2.298 2.367 89.4 93.0 93.5 92.2 94.9 2- 2 2.493 2.246 2.338 2.332 2.322 2.355 90.1 93.8 93.5 93.1 94.5 2- 3 2.493 2.203 2.325 2.336 2.317 2.351 88.4 93.3 93.7 92.9 94.3 AV G 89.3 93.3 93.6 92.8 94.6 3- 1 2.493 2.261 2.363 2.363 2.353 2.379 90.7 94.8 94.8 94.4 95.4 3- 2 2.493 2.185 2.296 2.330 2.317 2.344 87.6 92.1 93.5 92.9 94.0 3- 3 2.493 2.218 2.324 2.308 2.294 2.335 89.0 93.2 92.6 92.0 93.7 AV G 89.1 93.4 93.6 93.1 94.4 238 TAB L E A.4 6 C o re Dat a fo r P r o j ect AL - 6 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .548 2.359 2.3 7 2 2 .385 2.39 6 2 .386 2.380 9 2 .6 93.1 9 3 . 6 94.0 9 3 . 6 93.4 1 - 2 2 .548 2.366 2.3 7 9 2 .364 2.37 6 2 .381 2.380 9 2 .9 93.4 9 2 . 8 93.2 9 3 . 4 93.4 1 - 3 2 .548 2.291 2.3 4 0 2 .339 2.34 9 2 .342 2.354 8 9 .9 91.8 9 1 . 8 92.2 9 1 . 9 92.4 Av g. 9 1 .8 92.8 9 2 . 7 93.2 9 3 . 0 93.1 St d. 1 . 63 0.82 0.9 0 0.93 0. 95 0.59 2 - 1 2 .530 2.333 2.3 6 2 2 .360 2.36 5 2 .365 2.376 9 2 .2 93.4 9 3 . 3 93.5 9 3 . 5 93.9 2 - 2 2 .530 2.342 2.3 4 6 2 .330 2.33 0 2 .376 2.384 9 2 .6 92.7 9 2 . 1 92.1 9 3 . 9 94.2 2 - 3 2 .530 2.294 2.3 7 7 2 .341 2.36 0 2 .366 2.389 9 0 .7 94.0 9 2 . 5 93.3 9 3 . 5 94.4 Av g. 9 1 .8 93.3 9 2 . 6 93.0 9 3 . 6 94.2 St d. 1 . 01 0.61 0.6 0 0.75 0. 24 0.26 3- 1 3- 2 3- 3 Av g. St d. Roadw a y Core - Gm b R oadw ay C o re - % G mm 239 TAB L E A.4 7 C o re Dat a fo r P r o j ect AR -1 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .437 2.237 2.2 65 2.262 2.29 4 2 .295 2.282 9 1.8 92.9 92 .8 94.1 94 .2 93.6 1 - 2 2 .437 2.283 2.2 85 2.308 2.29 2 2 .306 2.299 9 3.7 93.8 94 .7 94.1 94 .6 94.3 1 - 3 2 .437 2.231 2.2 58 2.257 2.27 9 2 .283 2.290 9 1.5 92.7 92 .6 93.5 93 .7 94.0 Av g. 9 2.3 93.1 93 .4 93.9 94 .2 94.0 St d. 1 .17 0.57 1.1 5 0.33 0. 47 0.35 2 - 1 2 .429 2.242 2.2 83 2.261 2.31 4 2 .313 2.317 9 2.3 94.0 93 .1 95.3 95 .2 95.4 2 - 2 2 .429 2.233 2.2 56 2.300 2.26 6 2 .283 2.273 9 1.9 92.9 94 .7 93.3 94 .0 93.6 2 - 3 2 .429 2.236 2.2 69 2.284 2.28 4 2 .293 2.293 9 2.1 93.4 94 .0 94.0 94 .4 94.4 Av g. 9 2.1 93.4 93 .9 94.2 94 .5 94.5 St d. 0 .19 0.56 0.8 1 1.00 0. 63 0.91 3 - 1 2 .436 2.233 2.2 64 2.269 2.29 2 2 .318 2.309 9 1.7 92.9 93 .1 94.1 95 .2 94.8 3 - 2 2 .436 2.229 2.2 47 2.255 2.28 4 2 .276 2.267 9 1.5 92.2 92 .6 93.8 93 .4 93.1 3 - 3 2 .436 2.231 2.2 73 2.276 2.29 9 2 .275 2.283 9 1.6 93.3 93 .4 94.4 93 .4 93.7 Av g. 9 1.6 92.8 93 .0 94.1 94 .0 93.9 St d. 0 .08 0.54 0.4 4 0.31 1. 01 0.87 Roadw a y Core - Gm b R oadw ay C ore - % G mm 240 TAB L E A.48 C o re Data for Proj e c t AR -2 Sa m p le G m m I n - P lac e 3- Mont h 6 - M o n t h 1 - Y ear 2 - Y ear 4 - Y ear I n - P lac e 3- Mo nth 6 - M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 464 2.2 0 2 2 .2 30 2.2 4 7 2 .2 62 2.2 6 8 2 .26 3 89 .4 90 .5 91. 2 9 1 . 8 9 2 . 0 9 1 . 8 1- 2 2 . 464 2.1 8 8 2 .2 53 2.2 5 6 2 .2 55 2.2 6 1 2 .26 2 88 .8 91 .4 91. 6 9 1 . 5 9 1 . 8 9 1 . 8 1- 3 2 . 464 2.2 2 7 2 .2 51 2.2 5 9 2 .2 64 2.2 8 0 2 .28 7 90 .4 91 .4 91. 7 9 1 . 9 9 2 . 5 9 2 . 8 Av g . 89 .5 91 .1 91. 5 9 1 . 7 9 2 . 1 9 2 . 2 Std . 0. 80 0. 52 0. 25 0. 19 0. 39 0. 57 2- 1 2 . 448 2.1 7 7 2 .2 20 2.2 2 4 2 .3 10 2.2 3 1 2 .24 7 88 .9 90 .7 90. 8 9 4 . 4 9 1 . 1 9 1 . 8 2- 2 2 . 448 2.1 7 4 2 .2 10 2.2 3 0 2 .1 67 2.2 3 5 2 .24 2 88 .8 90 .3 91. 1 8 8 . 5 9 1 . 3 9 1 . 6 2- 3 2 . 448 2.2 0 1 2 .2 34 2.2 4 6 2 .2 67 2.2 5 7 2 .26 9 89 .9 91 .3 91. 7 9 2 . 6 9 2 . 2 9 2 . 7 Av g . 89 .2 90 .7 91. 2 9 1 . 8 9 1 . 5 9 2 . 0 Std . 0. 60 0. 49 0. 46 3. 00 0. 57 0. 59 3- 1 3- 2 3- 3 Av g . Std . Roadw a y Core - Gm b R oadw ay Core - % G mm 241 TAB L E A.49 C o re Data for Proj e c t AR -3 Sa m p le G m m I n- P lac e 3- Mont h 6- M o nt h 1 - Y ear 2 - Y ear 4 - Y ear I n- P lac e 3- Mo nth 6- M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 426 2.2 1 3 2 .2 79 2.2 7 9 2 .2 87 2.2 7 0 2 .23 8 91 .2 93 .9 93. 9 9 4 . 3 9 3 . 6 9 2 . 3 1- 2 2 . 426 2.2 2 1 2 .2 86 2.2 9 7 2 .2 85 2.2 9 1 2 .25 6 91 .5 94 .2 94. 7 9 4 . 2 9 4 . 4 9 3 . 0 1- 3 2 . 426 2.2 3 3 2 .3 21 2.3 2 7 2 .3 20 2.2 9 7 2 .32 9 92 .0 95 .7 95. 9 9 5 . 6 9 4 . 7 9 6 . 0 Av g . 91 .6 94 .6 94. 8 9 4 . 7 9 4 . 2 9 3 . 7 Std . 0. 41 0. 93 1. 00 0. 81 0. 58 1. 99 2- 1 2 . 436 2.2 1 4 2 .2 86 2.2 9 8 2 .3 04 2.3 0 9 2 .31 0 90 .9 93 .8 94. 3 9 4 . 6 9 4 . 8 9 4 . 8 2- 2 2 . 436 2.2 4 9 2 .3 34 2.3 2 6 2 .3 27 2.3 3 7 2 .33 8 92 .3 95 .8 95. 5 9 5 . 5 9 5 . 9 9 6 . 0 2- 3 2 . 436 2.2 2 1 2 .2 93 2.3 0 2 2 .3 06 2.3 1 0 2 .32 2 91 .2 94 .1 94. 5 9 4 . 7 9 4 . 8 9 5 . 3 Av g . 91 .5 94 .6 94. 8 9 4 . 9 9 5 . 2 9 5 . 4 Std . 0. 76 1. 06 0. 62 0. 52 0. 65 0. 58 3- 1 3- 2 3- 3 Av g . Std . Roadw a y Core - Gm b R oadw ay Core - % G mm 242 TAB L E A.50 C o re Data for Proj e c t AR -4 Sa m p le G m m I n - P lac e 3- Mont h 6 - M o n t h 1 - Y ear 2 - Y ear 4 - Y ear I n - P lac e 3- Mo nth 6 - M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 409 2.1 7 9 2 .2 55 2.2 4 0 2 .2 68 2.2 5 7 2 .26 3 90 .5 93 .6 93. 0 9 4 . 1 9 3 . 7 9 3 . 9 1- 2 2 . 409 2.1 9 1 2 .2 78 2.2 5 6 2 .2 78 2.2 7 2 2 .27 3 91 .0 94 .6 93. 6 9 4 . 6 9 4 . 3 9 4 . 4 1- 3 2 . 409 2.1 8 8 2 .2 52 2.2 2 9 2 .2 51 2.2 7 0 2 .26 7 90 .8 93 .5 92. 5 9 3 . 4 9 4 . 2 9 4 . 1 Av g . 90 .7 93 .9 93. 1 9 4 . 1 9 4 . 1 9 4 . 1 Std . 0. 26 0. 59 0. 56 0. 57 0. 34 0. 21 2- 1 2 . 392 2.1 5 9 2 .2 53 2.2 3 6 2 .2 48 2.2 6 1 2 .27 4 90 .3 94 .2 93. 5 9 4 . 0 9 4 . 5 9 5 . 1 2- 2 2 . 392 2.1 7 0 2 .2 68 2.2 4 2 2 .2 76 2.2 6 1 2 .27 6 90 .7 94 .8 93. 7 9 5 . 2 9 4 . 5 9 5 . 2 2- 3 2 . 392 2.2 1 2 2 .2 89 2.2 6 0 2 .2 86 2.2 8 1 2 .27 7 92 .5 95 .7 94. 5 9 5 . 6 9 5 . 4 9 5 . 2 Av g . 91 .2 94 .9 93. 9 9 4 . 9 9 4 . 8 9 5 . 1 Std . 1. 17 0. 76 0. 52 0. 82 0. 48 0. 06 3- 1 2 . 401 2.1 8 5 2 .2 52 0.0 0 0 2 .2 64 2.2 6 4 2 .27 6 91 .0 93 .8 0 . 0 9 4 . 3 9 4 . 3 9 4 . 8 3- 2 2 . 401 2.1 7 9 2 .2 50 0.0 0 0 2 .2 60 2.2 6 8 2 .27 4 90 .8 93 .7 0 . 0 9 4 . 1 9 4 . 5 9 4 . 7 3- 3 2 . 401 2.1 8 7 2 .2 60 0.0 0 0 2 .2 67 2.2 7 9 2 .27 6 91 .1 94 .1 0 . 0 9 4 . 4 9 4 . 9 9 4 . 8 Av g . 9 0 . 9 9 3 . 9 0. 0 9 4. 3 9 4. 6 9 4. 8 Std . 0. 17 0. 22 0. 00 0. 15 0. 32 0. 05 Roadw a y Core - Gm b R oadw ay Core - % G mm 243 TABL E A.51 Co re Data fo r Pro j ect C O -1 Sa mp le G m m I n- P la c e 3- Mo nt h 6 - M o nt h 1 - Y e ar 2- Y ea r 4- Y ear I n- P la c e 3- Mo nt h 6- M ont h 1- Y ear 2 - Y e ar 4- Y ea r 1- 1 2. 45 1 2. 255 2 . 3 29 2. 34 6 2. 370 2 . 3 99 2. 3 78 9 2. 0 95 .0 9 5. 7 96. 7 9 7. 9 97 . 0 1- 2 2. 45 1 2. 236 2 . 3 10 2. 33 4 2. 364 2 . 3 88 2. 3 48 9 1. 2 94 .2 9 5. 2 96. 5 9 7. 4 95 . 8 1- 3 2. 45 1 2. 238 2 . 4 47 2. 28 8 2. 329 2 . 3 58 2. 3 77 9 1. 3 99 .8 9 3. 3 95. 0 9 6. 2 97 . 0 Avg. 9 1. 5 96 .4 9 4. 8 96. 1 9 7. 2 96 . 6 St d. 0. 43 3 . 0 3 1. 25 0 . 9 0 0. 87 0 . 7 0 2- 1 2. 43 6 2. 341 2 . 3 82 2. 37 9 2. 389 2 . 4 12 2. 4 08 9 6. 1 97 .8 9 7. 7 98. 1 9 9. 0 98 . 9 2- 2 2. 43 6 2. 316 2 . 3 72 2. 36 5 2. 379 2 . 3 97 2. 3 92 9 5. 1 97 .4 9 7. 1 97. 7 9 8. 4 98 . 2 2- 3 2. 43 6 2. 280 2 . 3 45 2. 34 7 2. 359 2 . 3 91 2. 3 96 9 3. 6 96 .3 9 6. 3 96. 8 9 8. 2 98 . 4 Avg. 9 4. 9 97 .1 9 7. 0 97. 5 9 8. 5 98 . 5 St d. 1. 26 0 . 7 9 0. 66 0 . 6 3 0. 44 0 . 3 4 3- 1 2. 45 0 2. 329 2 . 3 92 2. 39 0 2. 386 2 . 4 13 2. 3 99 9 5. 1 97 .6 9 7. 6 97. 4 9 8. 5 97 . 9 3- 2 2. 45 0 2. 330 2 . 3 70 2. 38 2 2. 402 2 . 4 21 2. 4 05 9 5. 1 96 .7 9 7. 2 98. 0 9 8. 8 98 . 2 3- 3 2. 45 0 2. 324 2 . 3 92 2. 40 1 2. 407 2 . 4 24 2. 4 09 9 4. 9 97 .6 9 8. 0 98. 2 9 8. 9 98 . 3 Avg. 9 5. 0 97 .3 9 7. 6 97. 9 9 8. 7 98 . 1 St d. 0. 13 0 . 5 2 0. 39 0 . 4 5 0. 23 0 . 2 1 Ro ad w a y C o re - Gmb R oadw ay C o re - % G m m 244 TAB L E A.5 2 C o re Dat a fo r P r o j ect C O -2 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .428 2.336 2.3 9 7 2 .387 2.38 4 2 .403 2.401 9 6 .2 98.7 9 8 . 3 98.2 9 9 . 0 98.9 1 - 2 2 .428 2.299 2.3 6 2 2 .366 2.37 8 2 .389 2.372 9 4 .7 97.3 9 7 . 4 97.9 9 8 . 4 97.7 1 - 3 2 .428 2.304 2.3 6 1 2 .350 2.36 4 2 .371 2.355 9 4 .9 97.2 9 6 . 8 97.4 9 7 . 7 97.0 Av g. 9 5 .3 97.7 9 7 . 5 97.8 9 8 . 3 97.9 St d. 0 . 83 0.84 0.7 6 0.42 0. 66 0.96 2 - 1 2 .449 2.326 2.3 7 4 2 .385 2.36 9 2 .356 2.363 9 5 .0 96.9 9 7 . 4 96.7 9 6 . 2 96.5 2 - 2 2 .449 2.320 2.3 4 8 2 .359 2.37 1 2 .377 2.359 9 4 .7 95.9 9 6 . 3 96.8 9 7 . 1 96.3 2 - 3 2 .449 2.302 2.3 5 6 2 .350 2.36 5 2 .369 2.369 9 4 .0 96.2 9 6 . 0 96.6 9 6 . 7 96.7 Av g. 9 4 .6 96.3 9 6 . 6 96.7 9 6 . 7 96.5 St d. 0 . 51 0.54 0.7 4 0.12 0. 43 0.21 3 - 1 2 .449 2.295 2.3 5 2 2 .336 2.34 4 2 .353 2.339 9 3 .7 96.0 9 5 . 4 95.7 9 6 . 1 95.5 3 - 2 2 .449 2.318 2.3 3 6 2 .351 2.35 8 2 .362 2.353 9 4 .7 95.4 9 6 . 0 96.3 9 6 . 4 96.1 3 - 3 2 .449 2.318 2.3 5 3 2 .342 2.35 9 2 .364 2.357 9 4 .7 96.1 9 5 . 6 96.3 9 6 . 5 96.2 Av g. 9 4 .3 95.8 9 5 . 7 96.1 9 6 . 4 95.9 St d. 0 . 54 0.39 0.3 1 0.34 0. 24 0.39 Roadw a y Core - Gm b R oadw ay C o re - % G mm 245 TAB L E A.5 3 C o re Dat a fo r P r o j ect C O -3 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .427 2.283 2.3 02 2.337 2.31 6 2 .322 2.316 9 4.1 94.8 96 . 3 95.4 95 . 7 95.4 1 - 2 2 .427 2.257 2.2 80 2.335 2.30 9 2 .313 2.308 9 3.0 93.9 96 . 2 95.1 95 . 3 95.1 1 - 3 2 .427 2.250 2.2 79 2.363 2.29 9 2 .293 2.294 9 2.7 93.9 97 . 4 94.7 94 . 5 94.5 Av g. 9 3.3 94.2 96 . 6 95.1 95 . 2 95.0 St d. 0 . 72 0.54 0.6 4 0.35 0. 61 0.46 2 - 1 2 .435 2.276 2.3 26 2.353 2.33 6 2 .342 2.349 9 3.5 95.5 96 . 6 95.9 96 . 2 96.5 2 - 2 2 .435 2.287 2.3 30 2.323 2.34 1 2 .349 2.345 9 3.9 95.7 95 . 4 96.1 96 . 5 96.3 2 - 3 2 .435 2.280 2.2 86 2.297 2.35 0 2 .334 2.354 9 3.6 93.9 94 . 3 96.5 95 . 9 96.7 Av g. 9 3.7 95.0 95 . 5 96.2 96 . 2 96.5 St d. 0 . 23 1.00 1.1 5 0.29 0. 31 0.19 3- 1 3- 2 3- 3 Av g. St d. Roadw a y Core - Gm b R oadw ay C ore - % G mm 246 TAB L E A.5 4 C o re Dat a fo r P r o j ect C O -4 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .501 2.350 2.3 8 2 2 .297 2.35 6 2 .348 2.357 9 4 .0 95.2 9 1 . 8 94.2 9 3 . 9 94.2 1 - 2 2 .501 2.352 2.2 7 4 2 .295 2.36 6 2 .362 2.335 9 4 .0 90.9 9 1 . 8 94.6 9 4 . 4 93.4 1 - 3 2 .501 2.375 2.3 6 3 2 .290 2.37 9 2 .348 2.386 9 5 .0 94.5 9 1 . 6 95.1 9 3 . 9 95.4 Av g. 9 4 .3 93.5 9 1 . 7 94.6 9 4 . 1 94.3 St d. 0 . 56 2.31 0.1 4 0.46 0. 32 1.02 2 - 1 2 .497 2.333 2.3 4 8 2 .324 2.35 1 2 .364 2.382 9 3 .4 94.0 9 3 . 1 94.2 9 4 . 7 95.4 2 - 2 2 .497 2.308 2.2 9 3 2 .340 2.33 4 2 .341 2.353 9 2 .4 91.8 9 3 . 7 93.5 9 3 . 8 94.2 2 - 3 2 .497 2.363 2.3 2 5 2 .326 2.34 6 2 .338 2.348 9 4 .6 93.1 9 3 . 2 94.0 9 3 . 6 94.0 Av g. 9 3 .5 93.0 9 3 . 3 93.9 9 4 . 0 94.6 St d. 1 . 10 1.11 0.3 5 0.35 0. 57 0.74 3 - 1 2 .510 2.348 2.3 4 2 2 .349 2.36 6 2 .362 2.376 9 3 .5 93.3 9 3 . 6 94.3 9 4 . 1 94.7 3 - 2 2 .510 2.343 2.3 5 5 2 .353 2.37 8 2 .388 2.375 9 3 .3 93.8 9 3 . 7 94.7 9 5 . 1 94.6 3 - 3 2 .510 2.329 2.3 3 6 2 .324 2.34 7 2 .374 2.360 9 2 .8 93.1 9 2 . 6 93.5 9 4 . 6 94.0 Av g. 9 3 .2 93.4 9 3 . 3 94.2 9 4 . 6 94.4 St d. 0 . 39 0.39 0.6 3 0.62 0. 52 0.36 Roadw a y Core - Gm b R oadw ay C o re - % G mm 247 TAB L E A.5 5 C o re Dat a fo r P r o j ect C O -5 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .451 2.289 2.3 35 2.331 2.34 9 2 .329 2.330 9 3.4 95.3 95 . 1 95.8 95 . 0 95.1 1 - 2 2 .451 2.249 2.3 08 2.314 2.32 5 2 .313 2.312 9 1.8 94.2 94 . 4 94.9 94 . 4 94.3 1 - 3 2 .451 2.272 2.3 23 2.303 2.31 5 2 .316 2.300 9 2.7 94.8 94 . 0 94.5 94 . 5 93.8 Av g. 9 2.6 94.7 94 . 5 95.0 94 . 6 94.4 St d. 0 . 82 0.55 0.5 8 0.71 0. 35 0.62 2 - 1 2 .462 2.244 2.2 81 2.299 2.30 4 2 .293 2.259 9 1.1 92.6 93 . 4 93.6 93 . 1 91.8 2 - 2 2 .462 2.247 2.2 89 2.295 2.30 9 2 .295 2.289 9 1.3 93.0 93 . 2 93.8 93 . 2 93.0 2 - 3 2 .462 2.253 2.2 92 2.290 2.30 5 2 .301 2.300 9 1.5 93.1 93 . 0 93.6 93 . 5 93.4 Av g. 9 1.3 92.9 93 . 2 93.7 93 . 3 92.7 St d. 0 . 19 0.23 0.1 8 0.11 0. 17 0.86 3 - 1 2 .462 2.238 2.2 90 2.293 2.30 4 2 .302 2.290 9 0.9 93.0 93 . 1 93.6 93 . 5 93.0 3 - 2 2 .462 2.239 2.2 92 2.295 2.31 3 2 .301 2.266 9 0.9 93.1 93 . 2 93.9 93 . 5 92.0 3 - 3 2 .462 2.245 2.2 93 2.301 2.31 0 2 .295 2.302 9 1.2 93.1 93 . 5 93.8 93 . 2 93.5 Av g. 9 1.0 93.1 93 . 3 93.8 93 . 4 92.9 St d. 0 . 15 0.06 0.1 7 0.19 0. 15 0.74 Roadw a y Core - Gm b R oadw ay C ore - % G mm 248 TAB L E A.56 C o re Data f o r Project F L -1 Samp le G m m R o a dw ay Co re - Gmb R oad w ay Co re - % G mm In-Pl a c e 3-Month 6 -Month 1 -Y ear 2-Y ear In- P l a c e 3-Month 6 -Month 1 - Y ear 2-Y ear 1- 1 2 .460 2.233 2.298 2.317 2.303 2.318 90.8 93.4 94.2 93.6 94.2 1- 2 2 .460 2.285 2.319 2.332 2.336 2.349 92.9 94.3 94.8 95.0 95.5 1- 3 2 .460 2.277 2.302 2.331 2.329 2.337 92.6 93.6 94.8 94.7 95.0 AV G 92.1 93.8 94.6 94.4 94.9 2- 1 2 .450 2.258 2.337 2.330 2.353 2.352 92.2 95.4 95.1 96.0 96.0 2- 2 2 .450 2.196 2.282 2.313 2.320 2.331 89.6 93.1 94.4 94.7 95.1 2- 3 2 .450 2.274 2.333 2.340 2.253 2.343 92.8 95.2 95.5 92.0 95.6 AV G 91.5 94.6 95.0 94.2 95.6 3- 1 3- 2 3- 3 AV G 249 TAB L E A.57 Core Dat a for P r oject GA-1 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .540 2.448 2.4 33 2.442 2.44 7 2 .462 2.447 9 6.4 95.8 96 . 1 96.3 96 . 9 96.3 1 - 2 2 .540 2.414 2.4 19 2.426 2.41 7 2 .444 2.450 9 5.0 95.2 95 . 5 95.2 96 . 2 96.5 1 - 3 2 .540 2.396 2.4 32 2.423 2.41 5 2 .435 2.422 9 4.3 95.7 95 . 4 95.1 95 . 9 95.4 Av g. 9 5.2 95.6 95 . 7 95.5 96 . 3 96.0 St d. 1 . 04 0.31 0.4 0 0.71 0. 54 0.61 2 - 1 2 .520 2.405 2.4 08 2.422 2.41 8 2 .451 2.444 9 5.4 95.6 96 . 1 96.0 97 . 3 97.0 2 - 2 2 .520 2.393 2.4 22 2.415 2.44 7 2 .450 2.441 9 5.0 96.1 95 . 8 97.1 97 . 2 96.9 2 - 3 2 .520 2.417 2.4 03 2.424 2.43 3 2 .438 2.442 9 5.9 95.4 96 . 2 96.5 96 . 7 96.9 Av g. 9 5.4 95.7 96 . 0 96.5 97 . 1 96.9 St d. 0 . 48 0.39 0.1 9 0.58 0. 29 0.06 3 - 1 2 .537 2.415 2.4 40 2.433 2.43 5 2 .442 2.431 9 5.2 96.2 95 . 9 96.0 96 . 3 95.8 3 - 2 2 .537 2.385 2.4 28 2.424 2.43 5 2 .423 2.432 9 4.0 95.7 95 . 5 96.0 95 . 5 95.9 3 - 3 2 .537 2.368 2.4 31 2.426 2.43 6 2 .443 2.449 9 3.3 95.8 95 . 6 96.0 96 . 3 96.5 Av g. 9 4.2 95.9 95 . 7 96.0 96 . 0 96.1 St d. 0 . 94 0.25 0.1 9 0.02 0. 44 0.40 Roadw a y Core - Gm b R oadw ay C ore - % G mm 250 TAB L E A.5 8 C o re Dat a fo r P r o j ect I L -1 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .502 2.284 2.3 5 0 2 .326 2.34 6 2 .345 2.360 9 1 .3 93.9 9 3 . 0 93.8 9 3 . 7 94.3 1 - 2 2 .502 2.255 2.3 5 7 2 .340 2.34 0 2 .355 2.356 9 0 .1 94.2 9 3 . 5 93.5 9 4 . 1 94.2 1 - 3 2 .502 2.249 2.3 2 0 2 .311 2.33 8 2 .359 2.348 8 9 .9 92.7 9 2 . 4 93.4 9 4 . 3 93.8 Av g. 9 0 .4 93.6 9 3 . 0 93.6 9 4 . 0 94.1 St d. 0 . 75 0.79 0.5 8 0.17 0. 29 0.24 2 - 1 2 .499 2.247 2.3 2 5 2 .327 2.35 0 2 .349 2.366 8 9 .9 93.0 9 3 . 1 94.0 9 4 . 0 94.7 2 - 2 2 .499 2.312 2.3 7 8 2 .377 2.36 9 2 .373 2.382 9 2 .5 95.2 9 5 . 1 94.8 9 5 . 0 95.3 2 - 3 2 .499 2.346 2.3 9 5 2 .407 2.40 5 2 .402 2.404 9 3 .9 95.8 9 6 . 3 96.2 9 6 . 1 96.2 Av g. 9 2 .1 94.7 9 4 . 9 95.0 9 5 . 0 95.4 St d. 2 . 01 1.46 1.6 2 1.12 1. 06 0.76 3 - 1 2 .491 2.249 2.3 2 2 2 .333 2.33 6 2 .328 2.351 9 0 .3 93.2 9 3 . 7 93.8 9 3 . 5 94.4 3 - 2 2 .491 2.235 2.3 2 6 2 .324 2.34 8 2 .355 2.346 8 9 .7 93.4 9 3 . 3 94.3 9 4 . 5 94.2 3 - 3 2 .491 2.280 2.3 3 5 2 .332 2.33 9 2 .353 2.354 9 1 .5 93.7 9 3 . 6 93.9 9 4 . 5 94.5 Av g. 9 0 .5 93.4 9 3 . 5 94.0 9 4 . 2 94.4 St d. 0 . 92 0.27 0.2 0 0.25 0. 60 0.16 Roadw a y Core - Gm b R oadw ay C o re - % G mm 251 TAB L E A.5 9 C o re Dat a fo r P r o j ect I L -2 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .446 2.247 2.3 13 2.305 2.29 2 2 .324 2.333 9 1.9 94.6 94 . 2 93.7 95 . 0 95.4 1 - 2 2 .446 2.246 2.3 00 2.297 2.31 0 2 .321 2.318 9 1.8 94.0 93 . 9 94.4 94 . 9 94.8 1 - 3 2 .446 2.255 2.2 95 2.301 2.31 2 2 .329 2.328 9 2.2 93.8 94 . 1 94.5 95 . 2 95.2 Av g. 9 2.0 94.1 94 . 1 94.2 95 . 0 95.1 St d. 0 . 20 0.38 0.1 6 0.45 0. 17 0.31 2 - 1 2 .428 2.206 2.2 72 2.260 2.30 3 2 .300 2.301 9 0.9 93.6 93 . 1 94.9 94 . 7 94.8 2 - 2 2 .428 2.223 2.3 02 2.298 2.28 1 2 .318 2.322 9 1.6 94.8 94 . 6 93.9 95 . 5 95.6 2 - 3 2 .428 2.239 2.2 91 2.296 2.29 9 2 .323 2.331 9 2.2 94.4 94 . 6 94.7 95 . 7 96.0 Av g. 9 1.5 94.2 94 . 1 94.5 95 . 3 95.5 St d. 0 . 68 0.63 0.8 8 0.48 0. 50 0.63 3 - 1 2 .433 2.242 2.2 98 2.298 2.30 1 2 .322 2.325 9 2.1 94.5 94 . 5 94.6 95 . 4 95.6 3 - 2 2 .433 2.214 2.2 90 2.281 2.29 1 2 .318 2.324 9 1.0 94.1 93 . 8 94.2 95 . 3 95.5 3 - 3 2 .433 2.208 2.2 85 2.271 2.27 8 2 .299 2.305 9 0.8 93.9 93 . 3 93.6 94 . 5 94.7 Av g. 9 1.3 94.2 93 . 8 94.1 95 . 1 95.3 St d. 0 . 75 0.27 0.5 6 0.47 0. 51 0.46 Roadw a y Core - Gm b R oadw ay C ore - % G mm 252 TAB L E A.6 0 C o re Dat a fo r P r o j ect I L -3 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .505 2.283 2.3 4 1 2 .332 2.34 8 2 .343 2.359 9 1 .1 93.5 9 3 . 1 93.7 9 3 . 5 94.2 1 - 2 2 .505 2.284 2.3 4 0 2 .334 2.34 5 2 .353 2.359 9 1 .2 93.4 9 3 . 2 93.6 9 3 . 9 94.2 1 - 3 2 .505 2.295 2.3 5 6 2 .344 2.35 1 2 .359 2.365 9 1 .6 94.1 9 3 . 6 93.9 9 4 . 2 94.4 Av g. 9 1 .3 93.6 9 3 . 3 93.7 9 3 . 9 94.3 St d. 0 . 27 0.36 0.2 6 0.12 0. 32 0.14 2 - 1 2 .493 2.321 2.3 6 5 2 .370 2.37 1 2 .376 2.371 9 3 .1 94.9 9 5 . 1 95.1 9 5 . 3 95.1 2 - 2 2 .493 2.286 2.3 3 5 2 .317 2.35 0 2 .354 2.365 9 1 .7 93.7 9 2 . 9 94.3 9 4 . 4 94.9 2 - 3 2 .493 2.294 2.3 4 2 2 .329 2.34 8 2 .337 2.350 9 2 .0 93.9 9 3 . 4 94.2 9 3 . 7 94.3 Av g. 9 2 .3 94.2 9 3 . 8 94.5 9 4 . 5 94.7 St d. 0 . 74 0.63 1.1 1 0.51 0. 78 0.43 3 - 1 2 .493 2.297 2.3 6 6 2 .363 2.35 7 2 .362 2.355 9 2 .1 94.9 9 4 . 8 94.5 9 4 . 7 94.5 3 - 2 2 .493 2.331 2.3 7 3 2 .357 2.36 9 2 .372 2.361 9 3 .5 95.2 9 4 . 5 95.0 9 5 . 1 94.7 3 - 3 2 .493 2.331 2.3 7 3 2 .356 2.37 7 2 .371 2.381 9 3 .5 95.2 9 4 . 5 95.3 9 5 . 1 95.5 Av g. 9 3 .0 95.1 9 4 . 6 95.0 9 5 . 0 94.9 St d. 0 . 79 0.16 0.1 5 0.40 0. 22 0.55 Roadw a y Core - Gm b R oadw ay C o re - % G mm 253 TAB L E A.6 1 C o re Dat a fo r P r o j ect I N -1 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .465 2.200 2.1 78 2.168 2.24 3 2 .257 2.310 8 9.2 88.4 88 . 0 91.0 91 . 6 93.7 1 - 2 2 .465 2.233 2.2 05 2.201 2.26 2 2 .295 2.291 9 0.6 89.5 89 . 3 91.8 93 . 1 92.9 1 - 3 2 .465 2.221 2.2 43 2.247 2.29 9 2 .309 2.260 9 0.1 91.0 91 . 2 93.3 93 . 7 91.7 Av g. 9 0.0 89.6 89 . 5 92.0 92 . 8 92.8 St d. 0 . 68 1.32 1.6 1 1.16 1. 09 1.02 2 - 1 2 .469 2.259 2.3 09 2.240 2.35 3 2 .354 2.308 9 1.5 93.5 90 . 7 95.3 95 . 3 93.5 2 - 2 2 .469 2.235 2.2 92 2.269 2.29 3 2 .298 2.322 9 0.5 92.8 91 . 9 92.9 93 . 1 94.0 2 - 3 2 .469 2.267 2.2 47 2.286 2.31 0 2 .333 2.381 9 1.8 91.0 92 . 6 93.6 94 . 5 96.4 Av g. 9 1.3 92.5 91 . 7 93.9 94 . 3 94.7 St d. 0 . 67 1.30 0.9 4 1.25 1. 15 1.57 3 - 1 2 .471 2.262 2.2 53 2.258 2.30 5 2 .320 2.354 9 1.5 91.2 91 . 4 93.3 93 . 9 95.3 3 - 2 2 .471 2.321 2.1 62 2.200 2.26 3 2 .297 2.350 9 3.9 87.5 89 . 0 91.6 93 . 0 95.1 3 - 3 2 .471 2.282 2.1 77 2.201 2.25 9 2 .318 2.329 9 2.4 88.1 89 . 1 91.4 93 . 8 94.3 Av g. 9 2.6 88.9 89 . 8 92.1 93 . 6 94.9 St d. 1 . 21 1.97 1.3 4 1.03 0. 52 0.54 Roadw a y Core - Gm b R oadw ay C ore - % G mm 254 TAB L E A.6 2 C o re Dat a fo r P r o j ect I N -2 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .684 2.471 2.4 3 4 2 .468 2.56 1 2 .555 2.570 9 2 .1 90.7 9 2 . 0 95.4 9 5 . 2 95.8 1 - 2 2 .684 2.368 2.4 0 6 2 .461 2.53 7 2 .497 2.500 8 8 .2 89.6 9 1 . 7 94.5 9 3 . 0 93.1 1 - 3 2 .684 2.395 2.4 0 4 2 .479 2.51 5 2 .510 2.492 8 9 .2 89.6 9 2 . 4 93.7 9 3 . 5 92.8 Av g. 8 9 .8 90.0 9 2 . 0 94.5 9 3 . 9 93.9 St d. 1 . 99 0.62 0.3 4 0.86 1. 13 1.60 2 - 1 2 .673 2.423 2.4 1 7 2 .437 2.53 3 2 .571 2.557 9 0 .6 90.4 9 1 . 2 94.8 9 6 . 2 95.7 2 - 2 2 .673 2.475 2.3 9 5 2 .423 2.53 8 2 .505 2.559 9 2 .6 89.6 9 0 . 6 94.9 9 3 . 7 95.7 2 - 3 2 .673 2.472 2.4 3 2 2 .447 2.51 9 2 .528 2.531 9 2 .5 91.0 9 1 . 5 94.2 9 4 . 6 94.7 Av g. 9 1 .9 90.3 9 1 . 1 94.7 9 4 . 8 95.4 St d. 1 . 09 0.70 0.4 5 0.37 1. 25 0.58 3 - 1 2 .698 2.496 2.4 9 7 2 .489 2.57 8 2 .546 2.584 9 2 .5 92.6 9 2 . 3 95.6 9 4 . 4 95.8 3 - 2 2 .698 2.519 2.4 9 1 2 .506 2.54 2 2 .512 2.558 9 3 .4 92.3 9 2 . 9 94.2 9 3 . 1 94.8 3 - 3 2 .698 2.470 2.4 4 6 2 .443 2.55 4 2 .516 2.553 9 1 .5 90.7 9 0 . 5 94.7 9 3 . 3 94.6 Av g. 9 2 .5 91.8 9 1 . 9 94.8 9 3 . 6 95.1 St d. 0 . 91 1.03 1.2 1 0.68 0. 69 0.62 Roadw a y Core - Gm b R oadw ay C o re - % G mm 255 TAB L E A.6 3 C o re Dat a fo r P r o j ect KS -1 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .435 2.130 2.1 90 2.203 2.22 7 2 .295 2.294 8 7.5 89.9 90 . 5 91.5 94 . 3 94.2 1 - 2 2 .435 2.140 2.2 29 2.254 2.31 9 2 .306 2.255 8 7.9 91.5 92 . 6 95.2 94 . 7 92.6 1 - 3 2 .435 2.203 2.2 29 2.242 2.27 3 2 .282 2.253 9 0.5 91.5 92 . 1 93.3 93 . 7 92.5 Av g. 8 8.6 91.0 91 . 7 93.3 94 . 2 93.1 St d. 1 . 63 0.92 1.1 0 1.89 0. 49 0.95 2 - 1 2 .421 2.192 2.2 43 2.246 2.28 7 2 .281 2.245 9 0.5 92.6 92 . 8 94.5 94 . 2 92.7 2 - 2 2 .421 2.195 2.2 09 2.234 2.28 7 2 .270 2.273 9 0.7 91.2 92 . 3 94.5 93 . 8 93.9 2 - 3 2 .421 2.214 2.2 29 2.251 2.30 8 2 .293 2.285 9 1.4 92.1 93 . 0 95.3 94 . 7 94.4 Av g. 9 0.9 92.0 92 . 7 94.8 94 . 2 93.7 St d. 0 . 49 0.71 0.3 6 0.50 0. 48 0.85 3 - 1 2 .413 2.225 2.1 96 2.216 2.23 5 2 .231 2.214 9 2.2 91.0 91 . 8 92.6 92 . 5 91.8 3 - 2 2 .413 2.183 2.1 81 2.232 2.25 2 2 .248 2.256 9 0.5 90.4 92 . 5 93.3 93 . 2 93.5 3 - 3 2 .413 2.114 2.1 74 2.208 2.21 4 2 .202 2.215 8 7.6 90.1 91 . 5 91.8 91 . 3 91.8 Av g. 9 0.1 90.5 92 . 0 92.6 92 . 3 92.3 St d. 2 . 32 0.47 0.5 0 0.79 0. 96 0.99 Roadw a y Core - Gm b R oadw ay C ore - % G mm 256 TAB L E A.64 C o re Data for Proj e c t KY-1 Sa mple Gmm I n - P lac e 3- Mont h 6 - M o n t h 1 - Y ear 2 - Y ear 4 - Y ear I n - P lac e 3- Mo nth 6 - M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 480 2.1 2 5 2 .1 56 2.1 4 7 2 .2 01 2.2 2 6 2 .10 4 85 .7 86 .9 86. 6 8 8 . 8 8 9 . 8 8 4 . 8 1- 2 2 . 480 2.1 5 8 2 .1 69 2.1 5 8 2 .1 99 2.1 9 8 2 .14 7 87 .0 87 .5 87. 0 8 8 . 7 8 8 . 6 8 6 . 6 1- 3 2 . 480 2.1 6 6 2 .1 75 2.1 6 0 2 .1 64 2.1 9 7 2 .12 2 87 .3 87 .7 87. 1 8 7 . 3 8 8 . 6 8 5 . 6 Avg. 86 .7 87 .4 86. 9 8 8 . 2 8 9 . 0 8 5 . 7 St d . 0. 88 0. 39 0. 28 0. 84 0. 66 0. 87 2- 1 2 . 453 2.0 2 5 2 .0 68 2.1 2 4 2 .1 53 2.1 7 1 2 .19 8 82 .6 84 .3 86. 6 8 7 . 8 8 8 . 5 8 9 . 6 2- 2 2 . 453 2.1 2 5 2 .1 53 2.1 3 7 2 .1 70 2.1 8 1 2 .18 6 86 .6 87 .8 87. 1 8 8 . 5 8 8 . 9 8 9 . 1 2- 3 2 . 453 2.0 5 9 2 .1 94 2.0 9 9 2 .0 90 2.1 2 6 2 .21 4 83 .9 89 .4 85. 6 8 5 . 2 8 6 . 7 9 0 . 3 Avg. 84 .4 87 .2 86. 4 8 7 . 1 8 8 . 0 8 9 . 7 St d . 2. 07 2. 62 0. 79 1. 72 1. 19 0. 57 3- 1 3- 2 3- 3 Avg. St d . Roadw a y Core - Gm b R oadw ay Core - % G mm 257 TAB L E A.65 C ore Data for Proj e c t KY-2 Sa mple Gmm I n- P lac e 3- Mont h 6- M o nt h 1 - Y ear 2 - Y ear 4 - Y ear I n- P lac e 3- Mo nth 6- M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 488 2.2 8 6 2 .3 10 2.3 0 5 2 .3 20 2.3 2 9 2 .33 6 91 .9 92 .8 92. 6 9 3 . 2 9 3 . 6 9 3 . 9 1- 2 2 . 488 2.2 8 8 2 .3 07 2.3 0 8 2 .3 30 2.3 3 0 2 .34 0 92 .0 92 .7 92. 8 9 3 . 6 9 3 . 6 9 4 . 1 1- 3 2 . 488 2.2 9 2 2 .3 12 2.3 1 0 2 .3 32 2.3 4 2 2 .34 2 92 .1 92 .9 92. 8 9 3 . 7 9 4 . 1 9 4 . 1 Avg. 92 .0 92 .8 92. 8 9 3 . 5 9 3 . 8 9 4 . 0 St d . 0. 12 0. 10 0. 10 0. 26 0. 29 0. 12 2- 1 2 . 470 2.2 9 2 2 .3 28 2.3 4 0 2 .3 37 2.3 4 4 2 .34 5 92 .8 94 .3 94. 7 9 4 . 6 9 4 . 9 9 4 . 9 2- 2 2 . 470 2.2 8 0 2 .3 18 2.3 3 0 2 .3 46 2.3 4 0 2 .34 7 92 .3 93 .8 94. 3 9 5 . 0 9 4 . 7 9 5 . 0 2- 3 2 . 470 2.2 7 0 2 .2 84 2.2 8 3 2 .3 05 2.3 1 3 2 .32 6 91 .9 92 .5 92. 4 9 3 . 3 9 3 . 6 9 4 . 2 Avg. 92 .3 93 .5 93. 8 9 4 . 3 9 4 . 4 9 4 . 7 St d . 0. 45 0. 93 1. 23 0. 87 0. 68 0. 47 3- 1 3- 2 3- 3 Avg. St d . Roadw a y Core - Gm b R oadw ay Core - % G mm 258 TAB L E A.66 C o re Data for Proj e c t KY-3 Sa mple Gmm I n - P lac e 3- Mont h 6 - M o n t h 1 - Y ear 2 - Y ear 4 - Y ear I n - P lac e 3- Mo nth 6 - M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 484 2.2 8 5 2 .2 95 2.3 2 4 2 .2 88 2.3 0 9 2 .33 3 92 .0 92 .4 93. 6 9 2 . 1 9 3 . 0 9 3 . 9 1- 2 2 . 484 2.2 4 7 2 .2 48 2.3 5 4 2 .2 90 2.3 0 7 2 .30 1 90 .5 90 .5 94. 8 9 2 . 2 9 2 . 9 9 2 . 6 1- 3 2 . 484 2.3 2 7 2 .2 79 2.3 3 2 2 .3 25 2.3 0 0 2 .27 6 93 .7 91 .7 93. 9 9 3 . 6 9 2 . 6 9 1 . 6 Avg. 92 .0 91 .5 94. 1 9 2 . 6 9 2 . 8 9 2 . 7 St d . 1. 61 0. 96 0. 63 0. 84 0. 19 1. 15 2- 1 2 . 481 2.2 7 4 2 .3 05 2.3 2 7 2 .3 45 2.3 1 3 2 .35 8 91 .7 92 .9 93. 8 9 4 . 5 9 3 . 2 9 5 . 0 2- 2 2 . 481 2.3 1 7 2 .3 45 2.2 6 1 2 .3 62 2.3 4 3 2 .35 1 93 .4 94 .5 91. 1 9 5 . 2 9 4 . 4 9 4 . 8 2- 3 2 . 481 2.2 3 8 2 .3 09 2.2 7 9 2 .3 48 2.3 3 9 2 .33 2 90 .2 93 .1 91. 9 9 4 . 6 9 4 . 3 9 4 . 0 Avg. 91 .8 93 .5 92. 3 9 4 . 8 9 4 . 0 9 4 . 6 St d . 1. 59 0. 89 1. 38 0. 37 0. 66 0. 54 3- 1 2 . 486 2.3 3 0 2 .3 66 2.3 4 8 2 .3 70 2.3 7 8 2 .38 1 93 .7 95 .2 94. 4 9 5 . 3 9 5 . 7 9 5 . 8 3- 2 2 . 486 2.3 4 2 2 .3 36 2.3 7 4 2 .3 90 2.4 0 8 2 .40 1 94 .2 94 .0 95. 5 9 6 . 1 9 6 . 9 9 6 . 6 3- 3 2 . 486 2.3 3 2 2 .3 22 2.3 5 1 2 .3 52 2.3 5 6 2 .36 5 93 .8 93 .4 94. 6 9 4 . 6 9 4 . 8 9 5 . 1 Avg. 93 .9 94 .2 94. 8 9 5 . 4 9 5 . 8 9 5 . 8 St d . 0. 26 0. 90 0. 57 0. 76 1. 05 0. 73 Roadw a y Core - Gm b R oadw ay Core - % G mm 259 TABL E A.67 Co re Data fo r Pro j ect M I -1 Sa mp le G m m I n- P la c e 3- Mo nt h 6 - M o nt h 1 - Y e ar 2- Y ea r 4- Y ear I n- P la c e 3- Mo nt h 6- M ont h 1- Y ear 2 - Y e ar 4- Y ea r 1- 1 2. 47 8 2. 263 2 . 2 85 2. 31 3 2. 324 2 . 3 49 2. 3 69 9 1. 3 92 .2 9 3. 3 93. 8 9 4. 8 95 . 6 1- 2 2. 47 8 2. 272 2 . 2 85 2. 29 2 2. 315 2 . 3 54 2. 3 41 9 1. 7 92 .2 9 2. 5 93. 4 9 5. 0 94 . 5 1- 3 2. 47 8 2. 271 2 . 2 75 2. 29 4 2. 308 2 . 3 47 2. 3 31 9 1. 6 91 .8 9 2. 6 93. 1 9 4. 7 94 . 1 Avg. 9 1. 6 92 .1 9 2. 8 93. 4 9 4. 8 94 . 7 St d. 0. 20 0 . 2 3 0. 47 0 . 3 2 0. 15 0 . 7 9 2- 1 2. 47 2 2. 278 2 . 2 79 2. 29 6 2. 297 2 . 3 50 2. 3 33 9 2. 2 92 .2 9 2. 9 92. 9 9 5. 1 94 . 4 2- 2 2. 47 2 2. 319 2 . 2 67 2. 28 8 2. 317 2 . 3 68 2. 3 41 9 3. 8 91 .7 9 2. 6 93. 7 9 5. 8 94 . 7 2- 3 2. 47 2 2. 240 2 . 2 71 2. 26 8 2. 305 2 . 3 41 2. 3 44 9 0. 6 91 .9 9 1. 7 93. 2 9 4. 7 94 . 8 Avg. 9 2. 2 91 .9 9 2. 4 93. 3 9 5. 2 94 . 6 St d. 1. 60 0 . 2 5 0. 58 0 . 4 1 0. 56 0 . 2 3 3- 1 2. 49 7 2. 244 2 . 3 10 2. 31 8 2. 332 2 . 3 59 2. 3 38 8 9. 9 92 .5 9 2. 8 93. 4 9 4. 5 93 . 6 3- 2 2. 49 7 2. 247 2 . 3 04 2. 33 2 2. 350 2 . 3 58 2. 3 43 9 0. 0 92 .3 9 3. 4 94. 1 9 4. 4 93 . 8 3- 3 2. 49 7 2. 266 2 . 2 91 2. 29 6 2. 317 2 . 3 43 2. 3 45 9 0. 7 91 .8 9 2. 0 92. 8 9 3. 8 93 . 9 Avg. 9 0. 2 92 .2 9 2. 7 93. 4 9 4. 2 93 . 8 St d. 0. 48 0 . 3 9 0. 73 0 . 6 6 0. 36 0 . 1 4 Ro ad w a y C o re - Gmb R oadw ay C o re - % G m m 260 TAB L E A.68 C o re Data for Proj e c t M I -2 Sa mple Gmm I n - P lac e 3- Mont h 6 - M o n t h 1 - Y ear 2 - Y ear 4 - Y ear I n - P lac e 3- Mo nth 6 - M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 446 2.2 9 2 2 .2 81 2.3 4 8 2 .3 48 2.3 7 8 2 .38 8 93 .7 93 .3 96. 0 9 6 . 0 9 7 . 2 9 7 . 6 1- 2 2 . 446 2.3 3 0 2 .3 08 2.3 5 2 2 .3 22 2.3 8 4 2 .38 9 95 .3 94 .4 96. 2 9 4 . 9 9 7 . 5 9 7 . 7 1- 3 2 . 446 2.2 8 3 2 .3 02 2.3 5 9 2 .3 73 2.3 7 3 2 .37 3 93 .3 94 .1 96. 4 9 7 . 0 9 7 . 0 9 7 . 0 Avg. 94 .1 93 .9 96. 2 9 6 . 0 9 7 . 2 9 7 . 4 St d . 1. 02 0. 58 0. 23 1. 04 0. 23 0. 37 2- 1 2 . 440 2.2 7 5 2 .3 80 2.3 6 4 2 .3 87 2.3 9 4 2 .40 1 93 .2 97 .5 96. 9 9 7 . 8 9 8 . 1 9 8 . 4 2- 2 2 . 440 2.2 9 1 2 .3 85 2.3 8 6 2 .3 99 2.3 9 1 2 .40 8 93 .9 97 .7 97. 8 9 8 . 3 9 8 . 0 9 8 . 7 2- 3 2 . 440 2.2 7 7 2 .3 66 2.3 8 7 2 .3 84 2.3 8 8 2 .39 7 93 .3 97 .0 97. 8 9 7 . 7 9 7 . 9 9 8 . 2 Avg. 93 .5 97 .4 97. 5 9 8 . 0 9 8 . 0 9 8 . 4 St d . 0. 36 0. 40 0. 53 0. 33 0. 12 0. 23 3- 1 2 . 458 2.2 4 6 2 .3 39 2.3 3 4 2 .3 74 2.3 5 4 2 .38 4 91 .4 95 .2 95. 0 9 6 . 6 9 5 . 8 9 7 . 0 3- 2 2 . 458 2.2 5 7 2 .3 14 2.3 1 4 2 .3 72 2.3 3 1 2 .35 5 91 .8 94 .1 94. 1 9 6 . 5 9 4 . 8 9 5 . 8 3- 3 2 . 458 2.2 5 6 2 .3 53 2.3 3 3 2 .3 76 2.3 3 4 0 .00 0 91 .8 95 .7 94. 9 9 6 . 7 9 5 . 0 0 . 0 Avg. 91 .7 95 .0 94. 7 9 6 . 6 9 5 . 2 9 6 . 4 St d . 0. 25 0. 80 0. 46 0. 08 0. 51 0. 83 Roadw a y Core - Gm b R oadw ay Core - % G mm 261 TAB L E A.69 C ore Data for Proj e c t M I -3 Sa mple Gmm I n- P lac e 3- Mont h 6- M o nt h 1 - Y ear 2 - Y ear 4 - Y ear I n- P lac e 3- Mo nth 6- M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 468 2.2 8 8 2 .3 04 2.3 3 5 0 .0 00 2.3 8 7 2 .39 6 92 .7 93 .4 94. 6 0 . 0 96 .7 97 .1 1- 2 2 . 468 2.3 3 6 2 .3 11 2.3 3 8 0 .0 00 2.3 7 5 2 .37 8 94 .7 93 .6 94. 7 0 . 0 96 .2 96 .4 1- 3 2 . 468 2.2 8 2 2 .3 08 2.3 2 4 0 .0 00 2.3 8 8 2 .39 2 92 .5 93 .5 94. 2 0 . 0 96 .8 96 .9 Avg. 93 .3 93 .5 94. 5 0 . 0 96 .6 96 .8 St d . 1. 20 0. 14 0. 30 0. 00 0. 29 0. 38 2- 1 2 . 466 2.2 9 1 2 .3 07 2.3 1 4 0 .0 00 2.3 6 2 2 .37 5 92 .9 93 .6 93. 8 0 . 0 95 .8 96 .3 2- 2 2 . 466 2.2 9 7 2 .3 28 2.3 4 7 0 .0 00 2.3 8 3 2 .40 6 93 .1 94 .4 95. 2 0 . 0 96 .6 97 .6 2- 3 2 . 466 2.2 7 8 2 .3 07 2.3 2 6 0 .0 00 2.3 8 3 2 .38 8 92 .4 93 .6 94. 3 0 . 0 96 .6 96 .8 Avg. 92 .8 93 .8 94. 4 0 . 0 96 .4 96 .9 St d . 0. 39 0. 49 0. 68 0. 00 0. 49 0. 63 3- 1 3- 2 3- 3 Avg. St d . Roadw a y Core - Gm b R oadw ay Core - % G mm 262 TAB L E A.70 C o re Data fo r P r o j ect M O -1 Samp le G m m R o a dw ay Co re - Gmb R oad w ay Co re - % G mm In-Pl a c e 3-Month 6 -Month 1 -Y ear 2-Y ear In- P l a c e 3-Month 6 -Month 1 - Y ear 2-Y ear 1- 1 2 .474 2.287 2.381 2.354 2.360 92.4 96.2 95.1 95.4 1- 2 2 .474 2.300 2.388 2.334 2.384 93.0 96.5 94.3 96.4 1- 3 2 .474 2.359 2.381 2.345 2.378 95.4 96.2 94.8 96.1 AV G 93.6 96.3 94.8 96.0 2- 1 2 .476 2.302 2.410 2.404 2.369 93.0 97.3 97.1 95.7 2- 2 2 .476 2.319 2.386 2.379 2.386 93.7 96.4 96.1 96.4 2- 3 2 .476 2.328 2.396 2.384 2.379 94.0 96.8 96.3 96.1 AV G 93.6 96.8 96.5 96.0 3- 1 2 .485 2.309 2.392 2.377 2.358 92.9 96.3 95.7 94.9 3- 2 2 .485 2.330 2.389 2.373 2.385 93.8 96.1 95.5 96.0 3- 3 2 .485 2.309 2.378 2.373 2.373 92.9 95.7 95.5 95.5 AV G 93.2 96.0 95.5 95.5 263 TAB L E A.7 1 C o re Dat a fo r P r o j ect MO- 2 Sample Gmm In-Pl ac e 3-Month 6- M o nth 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .360 2.166 2.2 23 2.163 2.19 6 2 .219 2.221 9 1.8 94.2 91 . 7 93.1 94 . 0 94.1 1 - 2 2 .360 2.189 2.1 77 2.178 2.21 9 2 .252 2.237 9 2.8 92.2 92 . 3 94.0 95 . 4 94.8 1 - 3 2 .360 2.217 2.2 39 2.249 2.26 0 2 .259 2.290 9 3.9 94.9 95 . 3 95.8 95 . 7 97.0 Av g. 9 2.8 93.8 93 . 1 94.3 95 . 1 95.3 St d. 1 . 08 1.36 1.9 5 1.37 0. 91 1.53 2 - 1 2 .376 2.176 2.2 44 2.179 2.24 4 2 .270 2.248 9 1.6 94.4 91 . 7 94.4 95 . 5 94.6 2 - 2 2 .376 2.182 2.2 62 2.190 2.24 3 2 .289 2.268 9 1.8 95.2 92 . 2 94.4 96 . 3 95.5 2 - 3 2 .376 2.182 2.2 52 2.181 2.24 2 2 .276 2.274 9 1.8 94.8 91 . 8 94.4 95 . 8 95.7 Av g. 9 1.8 94.8 91 . 9 94.4 95 . 9 95.3 St d. 0 . 15 0.38 0.2 5 0.04 0. 41 0.57 3 - 1 2 .360 2.194 2.2 14 2.205 2.24 2 2 .234 2.226 9 3.0 93.8 93 . 4 95.0 94 . 7 94.3 3 - 2 2 .360 2.215 2.2 23 2.182 2.21 4 2 .230 2.219 9 3.9 94.2 92 . 5 93.8 94 . 5 94.0 3 - 3 2 .360 2.201 2.2 24 2.197 2.23 9 2 .212 2.244 9 3.3 94.2 93 . 1 94.9 93 . 7 95.1 Av g. 9 3.4 94.1 93 . 0 94.6 94 . 3 94.5 St d. 0 . 45 0.23 0.4 9 0.65 0. 50 0.55 Roadw a y Core - Gm b R oadw ay C ore - % G mm 264 TAB L E A.7 2 C o re Dat a fo r P r o j ect MO- 3 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .444 2.271 2.2 8 7 2 .284 2.31 9 2 .312 2.318 9 2 .9 93.6 9 3 . 5 94.9 9 4 . 6 94.8 1 - 2 2 .444 2.298 2.2 9 9 2 .281 2.33 4 2 .316 2.311 9 4 .0 94.1 9 3 . 3 95.5 9 4 . 8 94.6 1 - 3 2 .444 2.288 2.2 9 1 2 .288 2.31 8 2 .336 2.335 9 3 .6 93.7 9 3 . 6 94.8 9 5 . 6 95.5 Av g. 9 3 .5 93.8 9 3 . 5 95.1 9 5 . 0 95.0 St d. 0 . 56 0.25 0.1 4 0.37 0. 53 0.51 2 - 1 2 .434 2.266 2.2 9 3 2 .303 2.33 1 2 .321 2.331 9 3 .1 94.2 9 4 . 6 95.8 9 5 . 4 95.8 2 - 2 2 .434 2.272 2.2 9 5 2 .292 2.30 6 2 .313 2.326 9 3 .3 94.3 9 4 . 2 94.7 9 5 . 0 95.6 2 - 3 2 .434 2.267 2.2 9 7 2 .300 2.32 2 2 .346 2.327 9 3 .1 94.4 9 4 . 5 95.4 9 6 . 4 95.6 Av g. 9 3 .2 94.3 9 4 . 4 95.3 9 5 . 6 95.6 St d. 0 . 13 0.08 0.2 3 0.52 0. 71 0.11 3 - 1 2 .436 2.286 2.3 2 0 2 .315 2.33 7 2 .346 2.344 9 3 .8 95.2 9 5 . 0 95.9 9 6 . 3 96.2 3 - 2 2 .436 2.294 2.3 1 3 2 .326 2.32 9 2 .339 2.335 9 4 .2 95.0 9 5 . 5 95.6 9 6 . 0 95.9 3 - 3 2 .436 2.281 2.3 1 3 2 .312 2.32 5 2 .337 2.323 9 3 .6 95.0 9 4 . 9 95.4 9 5 . 9 95.4 Av g. 9 3 .9 95.0 9 5 . 1 95.7 9 6 . 1 95.8 St d. 0 . 27 0.17 0.3 0 0.25 0. 19 0.43 Roadw a y Core - Gm b R oadw ay C o re - % G mm 265 TABL E A. 73 C o re Dat a for P r oj ect NC -1 Sa mpl e G m m In-Pl a c e 3 - Month 6 - Month 1 -Y ear 2 - Y ear 4- Y ear In-Pl a c e 3 - Month 6 - Month 1 -Y ear 2-Y ear 4 - Y ear 1-1 2.640 2.418 2.469 2 . 407 2.4 54 2.458 2.492 91.6 93.5 91.2 93.0 93 . 1 94.4 1-2 2.640 2.396 2.448 2 . 417 2.4 57 2.472 2.472 90.8 92.7 91.6 93.1 93 . 6 93.6 1-3 2.640 2.330 2.414 2 . 403 2.4 22 2.433 2.458 88.3 91.4 91.0 91.7 92 . 2 93.1 Avg. 90.2 92.6 91.3 92.6 9 3 . 0 93.7 St d . 1.73 1.05 0.27 0.73 0. 75 0.65 2-1 2.638 2.416 2.471 2 . 460 2.4 65 2.471 2.481 91.6 93.7 93.3 93.4 93 . 7 94.0 2-2 2.638 2.350 2.435 2 . 429 2.4 37 2.465 2.484 89.1 92.3 92.1 92.4 93 . 4 94.2 2-3 2.638 2.363 2.431 2 . 397 2.4 43 2.449 2.458 89.6 92.2 90.9 92.6 92 . 8 93.2 Avg. 90.1 92.7 92.1 92.8 9 3 . 3 93.8 St d . 1.33 0.84 1.19 0.56 0. 43 0.54 3-1 2.649 2.374 2.460 2 . 418 2.4 73 2.489 2.498 89.6 92.9 91.3 93.4 94 . 0 94.3 3-2 2.649 2.381 2.463 2 . 443 2.4 86 2.489 2.498 89.9 93.0 92.2 93.8 94 . 0 94.3 3-3 2.649 2.401 2.466 2 . 445 2.4 79 2.480 2.484 90.6 93.1 92.3 93.6 93 . 6 93.8 Avg. 90.0 93.0 91.9 93.6 9 3 . 8 94.1 St d . 0.53 0.11 0.57 0.25 0. 20 0.31 Ro adw a y Co re - Gmb R oadw ay C ore - % G mm 266 TAB L E A.74 C o re Data for Proj e c t NE-1 Sa mple Gmm I n - P lac e 3- Mont h 6 - M o n t h 1 - Y ear 2 - Y ear 4 - Y ear I n - P lac e 3- Mo nth 6 - M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 414 0.0 0 0 2 .3 18 2.3 2 7 2 .3 27 2.3 2 5 2 .31 1 0. 0 9 6 . 0 96. 4 9 6 . 4 9 6 . 3 9 5 . 7 1- 2 2 . 414 0.0 0 0 2 .3 29 2.2 8 0 2 .3 23 2.3 2 2 2 .32 7 0. 0 9 6 . 5 94. 4 9 6 . 2 9 6 . 2 9 6 . 4 1- 3 2 . 414 2.2 3 4 2 .3 17 2.2 8 7 2 .3 28 2.3 3 7 2 .31 8 92 .5 96 .0 94. 7 9 6 . 4 9 6 . 8 9 6 . 0 Avg. 92 .5 96 .2 95. 2 9 6 . 4 9 6 . 4 9 6 . 1 St d . 0. 00 0. 28 1. 05 0. 11 0. 33 0. 33 2- 1 2 . 405 2.2 5 1 2 .2 74 2.2 6 9 2 .2 60 2.2 8 0 2 .27 8 93 .6 94 .6 94. 3 9 4 . 0 9 4 . 8 9 4 . 7 2- 2 2 . 405 2.2 0 5 2 .2 71 2.3 2 6 2 .2 80 2.2 8 6 2 .28 3 91 .7 94 .4 96. 7 9 4 . 8 9 5 . 1 9 4 . 9 2- 3 2 . 405 2.2 2 7 2 .2 81 2.3 2 4 2 .2 61 2.2 8 6 2 .29 1 92 .6 94 .8 96. 6 9 4 . 0 9 5 . 1 9 5 . 3 Avg. 92 .6 94 .6 95. 9 9 4 . 3 9 5 . 0 9 5 . 0 St d . 0. 96 0. 21 1. 34 0. 47 0. 14 0. 27 3- 1 3- 2 3- 3 Avg. St d . Roadw a y Core - Gm b R oadw ay Core - % G mm 267 TAB L E A.75 C ore Data for Proj e c t NE-2 Sa mple Gmm I n- P lac e 3- Mont h 6- M o nt h 1 - Y ear 2 - Y ear 4 - Y ear I n- P lac e 3- Mo nth 6- M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 437 2.2 5 6 2 .3 09 2.3 2 6 2 .3 26 2.3 4 1 2 .34 8 92 .6 94 .7 95. 4 9 5 . 4 9 6 . 1 9 6 . 3 1- 2 2 . 437 2.2 8 2 2 .3 19 2.3 1 0 2 .3 03 2.3 5 1 2 .33 8 93 .6 95 .2 94. 8 9 4 . 5 9 6 . 5 9 5 . 9 1- 3 2 . 437 2.2 8 5 2 .3 17 2.3 1 0 2 .3 28 2.3 2 6 2 .33 5 93 .8 95 .1 94. 8 9 5 . 5 9 5 . 4 9 5 . 8 Avg. 93 .3 95 .0 95. 0 9 5 . 2 9 6 . 0 9 6 . 0 St d . 0. 65 0. 22 0. 38 0. 57 0. 52 0. 28 2- 1 2 . 437 2.2 6 2 2 .3 20 2.3 2 2 2 .3 29 2.3 3 4 2 .33 7 92 .8 95 .2 95. 3 9 5 . 6 9 5 . 8 9 5 . 9 2- 2 2 . 437 2.2 6 1 2 .3 17 2.3 2 5 2 .3 35 2.3 3 9 2 .34 0 92 .8 95 .1 95. 4 9 5 . 8 9 6 . 0 9 6 . 0 2- 3 2 . 437 2.2 5 6 2 .3 18 2.3 2 9 2 .3 24 2.3 3 4 2 .33 4 92 .6 95 .1 95. 6 9 5 . 4 9 5 . 8 9 5 . 8 Avg. 92 .7 95 .1 95. 4 9 5 . 6 9 5 . 8 9 5 . 9 St d . 0. 13 0. 06 0. 14 0. 23 0. 12 0. 12 3- 1 2 . 443 2.2 7 0 2 .3 38 2.3 0 0 2 .3 36 2.3 3 2 2 .34 0 92 .9 95 .7 94. 1 9 5 . 6 9 5 . 5 9 5 . 8 3- 2 2 . 443 2.2 7 9 2 .3 40 2.3 2 5 2 .3 43 2.3 1 8 2 .33 2 93 .3 95 .8 95. 2 9 5 . 9 9 4 . 9 9 5 . 5 3- 3 2 . 443 2.2 6 3 2 .3 26 2.3 1 1 2 .2 99 2.3 3 6 2 .34 0 92 .6 95 .2 94. 6 9 4 . 1 9 5 . 6 9 5 . 8 Avg. 92 .9 95 .6 94. 6 9 5 . 2 9 5 . 3 9 5 . 7 St d . 0. 33 0. 31 0. 51 0. 97 0. 39 0. 19 Roadw a y Core - Gm b R oadw ay Core - % G mm 268 TAB L E A.76 C o re Data for Proj e c t NE-3 Sa mple Gmm I n - P lac e 3- Mont h 6 - M o n t h 1 - Y ear 2 - Y ear 4 - Y ear I n - P lac e 3- Mo nth 6 - M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 405 2.1 9 1 2 .3 00 2.2 8 9 2 .3 00 2.2 9 2 2 .31 4 91 .1 95 .6 95. 2 9 5 . 6 9 5 . 3 9 6 . 2 1- 2 2 . 405 2.1 7 1 2 .2 74 2.2 8 8 2 .2 96 2.3 0 1 2 .23 9 90 .3 94 .6 95. 1 9 5 . 5 9 5 . 7 9 3 . 1 1- 3 2 . 405 2.1 7 4 2 .2 53 2.2 6 8 2 .2 69 2.2 8 8 2 .29 2 90 .4 93 .7 94. 3 9 4 . 3 9 5 . 1 9 5 . 3 Avg. 90 .6 94 .6 94. 9 9 5 . 1 9 5 . 4 9 4 . 9 St d . 0. 45 0. 98 0. 49 0. 70 0. 28 1. 60 2- 1 2 . 390 2.2 1 9 2 .2 80 2.2 8 5 2 .2 81 2.2 9 1 2 .29 8 92 .8 95 .4 95. 6 9 5 . 4 9 5 . 9 9 6 . 2 2- 2 2 . 390 2.1 6 5 2 .2 74 2.2 8 1 2 .2 85 2.2 8 9 2 .29 3 90 .6 95 .1 95. 4 9 5 . 6 9 5 . 8 9 5 . 9 2- 3 2 . 390 2.1 5 8 2 .2 70 2.2 8 1 2 .2 69 2.2 8 6 2 .25 9 90 .3 95 .0 95. 4 9 4 . 9 9 5 . 6 9 4 . 5 Avg. 91 .2 95 .2 95. 5 9 5 . 3 9 5 . 8 9 5 . 5 St d . 1. 40 0. 21 0. 10 0. 35 0. 11 0. 89 3- 1 2 . 398 2.1 8 8 2 .2 68 2.2 8 3 2 .2 70 2.2 8 8 2 .29 1 91 .2 94 .6 95. 2 9 4 . 7 9 5 . 4 9 5 . 5 3- 2 2 . 398 2.2 0 1 2 .2 68 2.2 6 8 2 .2 62 2.2 7 5 2 .27 4 91 .8 94 .6 94. 6 9 4 . 3 9 4 . 9 9 4 . 8 3- 3 2 . 398 2.1 6 4 2 .2 67 2.2 7 1 2 .2 60 2.2 6 7 2 .27 9 90 .2 94 .5 94. 7 9 4 . 2 9 4 . 5 9 5 . 0 Avg. 91 .1 94 .6 94. 8 9 4 . 4 9 4 . 9 9 5 . 1 St d . 0. 78 0. 02 0. 33 0. 22 0. 44 0. 36 Roadw a y Core - Gm b R oadw ay Core - % G mm 269 TAB L E A.77 C ore Data for Proj e c t NE-4 Sa mple Gmm I n- P lac e 3- Mont h 6- M o nt h 1 - Y ear 2 - Y ear 4 - Y ear I n- P lac e 3- Mo nth 6- M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2 . 444 2.2 5 1 2 .3 10 2.3 1 4 2 .3 55 2.3 7 3 2 .38 4 92 .1 94 .5 94. 7 9 6 . 4 9 7 . 1 9 7 . 5 1- 2 2 . 444 2.2 5 3 2 .3 17 2.3 1 9 2 .3 66 2.3 6 4 2 .37 1 92 .2 94 .8 94. 9 9 6 . 8 9 6 . 7 9 7 . 0 1- 3 2 . 444 2.2 6 0 2 .3 34 2.3 3 5 2 .3 74 2.3 8 1 2 .38 4 92 .5 95 .5 95. 5 9 7 . 1 9 7 . 4 9 7 . 5 Avg. 92 .3 94 .9 95. 0 9 6 . 8 9 7 . 1 9 7 . 4 St d . 0. 19 0. 51 0. 45 0. 39 0. 35 0. 31 2- 1 2 . 438 2.2 4 0 2 .3 11 2.3 2 5 2 .3 56 2.3 7 5 2 .37 7 91 .9 94 .8 95. 4 9 6 . 6 9 7 . 4 9 7 . 5 2- 2 2 . 438 2.2 5 6 2 .3 19 2.3 1 7 2 .3 59 2.3 7 9 2 .38 1 92 .5 95 .1 95. 0 9 6 . 8 9 7 . 6 9 7 . 7 2- 3 2 . 438 2.2 6 4 2 .3 32 2.3 4 1 2 .3 71 2.3 7 2 2 .38 9 92 .9 95 .7 96. 0 9 7 . 3 9 7 . 3 9 8 . 0 Avg. 92 .4 95 .2 95. 5 9 6 . 9 9 7 . 4 9 7 . 7 St d . 0. 50 0. 43 0. 50 0. 33 0. 14 0. 25 3- 1 2 . 449 2.2 4 3 2 .3 11 2.3 2 4 2 .3 61 2.3 8 3 2 .38 3 91 .6 94 .4 94. 9 9 6 . 4 9 7 . 3 9 7 . 3 3- 2 2 . 449 2.2 4 0 2 .3 18 2.3 2 9 2 .3 58 2.3 8 1 2 .38 0 91 .5 94 .7 95. 1 9 6 . 3 9 7 . 2 9 7 . 2 3- 3 2 . 449 2.2 6 5 2 .3 15 2.3 2 4 2 .3 67 2.3 6 6 2 .37 2 92 .5 94 .5 94. 9 9 6 . 7 9 6 . 6 9 6 . 9 Avg. 91 .8 94 .5 95. 0 9 6 . 4 9 7 . 0 9 7 . 1 St d . 0. 56 0. 14 0. 12 0. 19 0. 38 0. 23 Roadw a y Core - Gm b R oadw ay Core - % G mm 270 TAB L E A.7 8 C o re Dat a fo r P r o j ect TN-1 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .459 2.231 2.2 8 9 2 .278 2.30 5 2 .316 2.298 9 0 .7 93.1 9 2 . 6 93.7 9 4 . 2 93.5 1 - 2 2 .459 2.212 2.2 8 4 2 .282 2.32 0 2 .312 2.293 9 0 .0 92.9 9 2 . 8 94.3 9 4 . 0 93.2 1 - 3 2 .459 2.189 2.2 7 6 2 .274 2.30 3 2 .308 2.303 8 9 .0 92.6 9 2 . 5 93.7 9 3 . 9 93.7 Av g. 8 9 .9 92.8 9 2 . 6 93.9 9 4 . 0 93.5 St d. 0 . 86 0.27 0.1 6 0.38 0. 16 0.20 2 - 1 2 .467 2.221 2.2 7 2 2 .266 2.30 5 2 .297 2.304 9 0 .0 92.1 9 1 . 9 93.4 9 3 . 1 93.4 2 - 2 2 .467 2.222 2.2 8 5 2 .301 2.31 5 2 .330 2.288 9 0 .1 92.6 9 3 . 3 93.8 9 4 . 4 92.7 2 - 3 2 .467 2.267 2.2 9 3 2 .298 2.31 1 2 .318 2.290 9 1 .9 92.9 9 3 . 1 93.7 9 4 . 0 92.8 Av g. 9 0 .7 92.6 9 2 . 8 93.6 9 3 . 8 93.0 St d. 1 . 07 0.43 0.7 9 0.20 0. 68 0.35 3 - 1 2 .464 2.295 2.3 1 2 2 .306 2.32 7 2 .351 2.327 9 3 .1 93.8 9 3 . 6 94.4 9 5 . 4 94.4 3 - 2 2 .464 2.294 2.3 1 8 2 .323 2.33 5 2 .355 2.330 9 3 .1 94.1 9 4 . 3 94.8 9 5 . 6 94.6 3 - 3 2 .464 2.263 2.3 1 0 2 .312 2.33 5 2 .317 2.309 9 1 .8 93.8 9 3 . 8 94.8 9 4 . 0 93.7 Av g. 9 2 .7 93.9 9 3 . 9 94.7 9 5 . 0 94.2 St d. 0 . 74 0.17 0.3 5 0.19 0. 85 0.46 Roadw a y Core - Gm b R oadw ay C o re - % G mm 271 TAB L E A.79 C ore Data for Proj e c t UT-1 Sa mple Gmm I n- P lac e 3- Mont h 6- M o nt h 1 - Y ear 2 - Y ear 4 - Y ear I n- P lac e 3- Mo nth 6- M ont h 1 - Y ear 2 - Y ear 4 - Y ear 1- 1 2. 470 2.2 69 2.3 31 2.3 27 2.3 17 2 . 31 3 91 . 9 94 . 4 94. 2 93 . 8 93 . 6 1- 2 2. 470 2.2 87 2.3 39 2.3 36 2.3 31 2 . 32 3 92 . 6 94 . 7 94. 6 94 . 4 94 . 0 1- 3 2. 470 2.2 46 2.3 14 2.3 20 2.3 10 2 . 29 5 90 . 9 93 . 7 93. 9 93 . 5 92 . 9 Avg. 91 .8 94 .3 94. 2 9 3 . 9 9 3 . 5 St d . 0. 83 0. 52 0. 32 0. 43 0. 57 2- 1 2. 458 2.3 10 2.3 10 2.3 16 2.2 97 2 . 30 2 94 . 0 94 . 0 94. 2 93 . 4 93 . 7 2- 2 2. 458 2.3 13 2.3 19 2.3 23 2.3 18 2 . 29 6 94 . 1 94 . 3 94. 5 94 . 3 93 . 4 2- 3 2. 458 2.2 70 2.3 23 2.2 96 2.3 09 2 . 32 8 92 . 4 94 . 5 93. 4 93 . 9 94 . 7 Avg. 93 .5 94 .3 94. 0 9 3 . 9 9 3 . 9 St d . 0. 98 0. 27 0. 57 0. 43 0. 69 3- 1 2. 465 2.2 20 2.2 11 2.2 24 2.3 15 2 . 30 2 90 . 1 89 . 7 90. 2 93 . 9 93 . 4 3- 2 2. 465 2.2 20 2.3 00 2.2 38 2.2 49 2 . 29 2 90 . 1 93 . 3 90. 8 91 . 2 93 . 0 3- 3 2. 465 2.2 44 2.2 97 2.2 87 2.3 26 2 . 29 8 91 . 0 93 . 2 92. 8 94 . 4 93 . 2 Avg. 90 .4 92 .1 91. 3 9 3 . 2 9 3 . 2 St d . 0. 56 2. 05 1. 34 1. 69 0. 20 Roadw a y Core - Gm b R oadw ay Core - % G mm 272 TAB L E A.8 0 C o re Dat a fo r P r o j ect W I -1 Sample Gmm In-Pl a c e 3-Month 6 - M o n th 1 - Y e ar 2- Y ear 4-Y ear In - P lace 3- Mo nth 6 - M on th 1-Y ear 2- Y ear 4-Y ear 1 - 1 2 .563 2.409 2.4 0 6 2 .412 2.42 1 2 .425 2.407 9 4 .0 93.9 9 4 . 1 94.5 9 4 . 6 93.9 1 - 2 2 .563 2.320 2.3 6 8 2 .392 2.40 2 2 .408 2.410 9 0 .5 92.4 9 3 . 3 93.7 9 4 . 0 94.0 1 - 3 2 .563 2.338 2.3 9 1 2 .377 2.38 9 2 .386 2.383 9 1 .2 93.3 9 2 . 7 93.2 9 3 . 1 93.0 Av g. 9 1 .9 93.2 9 3 . 4 93.8 9 3 . 9 93.6 St d. 1 . 84 0.75 0.6 9 0.63 0. 76 0.58 2 - 1 2 .558 2.408 2.4 2 7 2 .448 2.43 4 2 .420 2.417 9 4 .1 94.9 9 5 . 7 95.2 9 4 . 6 94.5 2 - 2 2 .558 2.397 2.4 3 4 2 .426 2.43 7 2 .432 2.430 9 3 .7 95.2 9 4 . 8 95.3 9 5 . 1 95.0 2 - 3 2 .558 2.367 2.4 1 6 2 .393 2.42 3 2 .405 2.421 9 2 .5 94.4 9 3 . 5 94.7 9 4 . 0 94.6 Av g. 9 3 .5 94.8 9 4 . 7 95.0 9 4 . 6 94.7 St d. 0 . 83 0.35 1.0 8 0.29 0. 53 0.26 3 - 1 2 .546 2.351 2.4 0 3 2 .398 2.41 4 2 .413 2.408 9 2 .3 94.4 9 4 . 2 94.8 9 4 . 8 94.6 3 - 2 2 .546 2.326 2.3 7 0 2 .363 2.40 3 2 .396 2.390 9 1 .4 93.1 9 2 . 8 94.4 9 4 . 1 93.9 3 - 3 2 .546 2.339 2.3 6 2 2 .359 2.39 3 2 .397 2.403 9 1 .9 92.8 9 2 . 7 94.0 9 4 . 1 94.4 Av g. 9 1 .9 93.4 9 3 . 2 94.4 9 4 . 3 94.3 St d. 0 . 49 0.85 0.8 4 0.41 0. 37 0.36 Roadw a y Core - Gm b R oadw ay C o re - % G mm