Georgia long-term pavement performance (GALTPP) program - maintaining Georgia's calibration sites and identifying the potential for using MEPDG for characterization of non-standard materials and methods (phase 1) [Oct. 2016]

GEORGIA DOT RESEARCH PROJECT 14-25 FINAL REPORT
GEORGIA LONG-TERM PAVEMENT PERFORMANCE (GALTPP) PROGRAM MAINTAINING GEORGIA'S CALIBRATION SITES AND IDENTIFYING THE POTENTIAL FOR USING MEPDG FOR CHARACTERIZATION OF NON-STANDARD MATERIALS AND METHODS (PHASE 1)
OFFICE OF RESEARCH 15 KENNEDY DRIVE
FOREST PARK, GA 30297-2534

1.Report No.: FHWA-GA-16-1425

2. Government Accession 3. Recipient's Catalog No.: No.:

4. Title and Subtitle: Georgia Long-Term Pavement Performance (GALTPP) Program Maintaining Georgia's Calibration Sites and Identifying The Potential for Using MEPDG For Characterization of Non-standard Materials and Methods (Phase 1)
7. Author(s): Yi-Ching Wu; Yichang (James) Tsai
9. Performing Organization Name and Address: Georgia Institute of Technology 790 Atlantic Drive Atlanta, GA 30332-0355

5. Report Date: October 2016 6. Performing Organization Code:
8. Performing Organ. Report No.: RP 14-25
10. Work Unit No.: 11. Contract or Grant No.:

12. Sponsoring Agency Name and Address:

13. Type of Report and Period Covered:

Georgia Department of Transportation

Final; March 2014 October 2016

Office of Research

14. Sponsoring Agency Code:

15 Kennedy Drive

Forest Park, GA 30297-2534

15. Supplementary Notes:

Prepared in cooperation with the U.S. Department of Transportation, Federal Highway Administration.

16. Abstract:

The Georgia Department of Transportation (GDOT) has initiated a Georgia Long-Term Pavement Performance

(GALTPP) monitoring program 1) to provide data for calibrating the prediction models in the AASHTO

Mechanistic-Empirical Pavement Design Guide (MEPDG) and 2) to monitor sites for evaluating the effect of

various materials and methods on pavement performance. A total of 38 flexible pavement sites (17 LTPP and 21

non-LTPP) and 23 rigid pavement sites (11 LTPP and 12 non-LTPP sites) sites were selected for the MEPDG

calibration and various field and laboratory testing, including condition surveys in accordance with LTPP Distress

Identification Manual, Falling Weight Deflectometer (FWD), etc., were conducted on the non-LTPP sites. The

detailed data collected for the calibration, is very valuable for future recalibration of the MEDPG and evaluation of

the performance of different pavement designs and materials. Thus, the objectives of this project are 1) to maintain

the data (e.g., distress data and FWD data) that has been collected on the sites, and 2) to identify the potential for

the characterization of non-standard methods and materials using the MEPDG to provide suggestions on the

implementation. In Phase 1 of this project, a GALTPP database was designed and populated with the inputs used

for initial MEPDG calibration and the data (e.g., distress data, FWD, etc.) collected on the non-LTPP sites with a

focus on flexible pavement sites. The site location data were corrected to ensure they can be correctly located for

long-term monitoring using core data collected by GDOT's coring data collection application. A GIS (Geographic

Information Systems) project, along with add-in tools, was developed to visualize the sites. In addition, the

predicted distresses by the MEPDG on interstate sites were verified and the differences in the designs between the

MEPDG and the 1972 AASHTO Interim Design Guide were discussed to provide suggestions on the MEPDG

implementation.

17. Key Words:

18. Distribution Statement:

Georgia LTPP; GALTPP

No Restriction

19. Security Classification (of this report): Unclassified

20. Security classification (of this page): Unclassified

21. Number of Pages: 22. Price: 84

Contract Research
GDOT Research Project No. 14-25 Draft Report
GEORGIA LONG-TERM PAVEMENT PERFORMANCE (GALTPP) PROGRAM MAINTAINING GEORGIA'S CALIBRATION SITES AND IDENTIFYING THE POTENTIAL FOR USING MEPDG FOR CHARACTERIZATION OF NON-STANDARD MATERIALS AND
METHODS (PHASE 1)
By Yi-Ching Wu, Research Engineer Yichang (James) Tsai, Ph.D., P.E.
Georgia Institute of Technology
Contract with Georgia Department of Transportation
In cooperation with U.S. Department of Transportation
Federal Highway Administration October 2016
The contents of this report reflect the views of the author(s) who is (are) responsible for the facts and the accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of the Georgia Department of Transportation or of the Federal Highway Administration. This report does not constitute a standard, specification, or regulation.

TABLE OF CONTENTS TABLE OF CONTENTS................................................................................................................ ii LIST OF TABLES ......................................................................................................................... iv LIST OF FIGURES ....................................................................................................................... vi EXECUTIVE SUMMARY ......................................................................................................... viii ACKNOWLEDGMENTS ........................................................................................................... xiv 1. INTRODUCTION................................................................................................................... 1
1. 1 Background and Research Need .......................................................................................... 1 1.2 Significance of Research....................................................................................................... 3 1.3 Research Objectives and Scope ............................................................................................ 4 1.4 Organization of this Report................................................................................................... 6 2. MANAGEMENT OF GALTPP DATA .................................................................................. 7 2.1 Data Gathered for the GALTPP Program ............................................................................. 7 2.2 Correction of Site Locations ............................................................................................... 10 2.3 Design of GALTPP Database ............................................................................................. 12 2.4 Visualization of GALTPP Data in a GIS Project................................................................ 16 3. PAVEMENT DESIGN, MAINTENANCE PRACTICES, AND PERFORMANCE OF
GEGORIA'S INSTERSTATE HIGHWAYS ....................................................................... 19 3.1 Pavement Design and Maintenance Practices..................................................................... 19 3.2 Predominant Distresses on Interstates ................................................................................ 22
ii

4. EVALUATION OF DESIGN OF GEORGIA'S INTERSTATE PAVEMENT STRUCTURE(S) USING DIFFERENT METHODS........................................................... 29
4.1 Review of MEPDG Predicted Distresses in Interstate Sites ............................................... 30 4.1.1 Observed Distresses ..................................................................................................... 32 4.1.2 Predicted Distresses ..................................................................................................... 35
4.2 Case Study on I-95 Site in Chatham County ...................................................................... 36 4.2.1 Analysis Using 1972 AASHTO Interim Design Guide ............................................... 37 4.2.2 Analysis Using MEPDG .............................................................................................. 38
4.3 Characterization of SMA .................................................................................................... 41 5. CONCLUSIONS AND RECOMMENDATAIONS............................................................. 45 REFERENCES ............................................................................................................................. 51 APPENDIX A: SITE LOCATION ............................................................................................. A-1 APPENDIX B: GALTPP TABLES............................................................................................ B-1 APPENDIX C: STUDIES RELATED TO SMA ....................................................................... C-1
iii

LIST OF TABLES TABLE 1.1 WORK BY PHASES .................................................................................................. 3 TABLE 2.1 EXAMPLE TEMPLATE FOR RECORDING DATA WITH LOCATION INFORMATION........................................................................................................................... 10 TABLE 3.1 DISTRESSES IN PACES ......................................................................................... 23 TABLE 4.1 GEORGIA'S CALIBRATION COEFFICIENTS .................................................... 30 TABLE 4.3 INTERSTATE SITES............................................................................................... 32
iv

v

LIST OF FIGURES
FIGURE 2.1 SITE AND CORE LOCATIONS............................................................................ 11 FIGURE 2.2 VERIFICATION OF SITE LOCATION (I-85 SITE) ............................................ 11 FIGURE 2.3 CONCEPTUAL DESIGN FOR GALTPP DATABASE........................................ 13 FIGURE 2.4 AN EXAMPLE OF VISUALIZING DATA IN GIS .............................................. 18 FIGURE 3.1 TYPICAL PAVEMENT DESIGN IN GEORGIA'S INTERSTATE HIGHWAYS ....................................................................................................................................................... 20 FIGURE 3.2 INTERSTATE LIFE-CYCLE MAINTENANCE ACTIVITIES............................ 22 FIGURE 3.3 INTERSTATE RATING DISTRIBUTION (FY 2015) .......................................... 24 FIGURE 3.4 INTERSTATE DISTRESSES FREQUENCY (FY 2015) ...................................... 25 FIGURE 3.5 PHOTO TAKEN ON I-95 NORTHBOUND MP 13-14 CHATHAM COUNTY IN 2014............................................................................................................................................... 26 FIGURE 3.6 PHOTOS TAKEN ON I-95 BEFORE AND AFTER RESURFACING ................ 27 FIGURE 4.1 GALTPP SITES ...................................................................................................... 31 FIGURE 4.2 MEASURED FATIGUE CRACKING ................................................................... 33 FIGURE 4.3 MEASURED RUTTING......................................................................................... 34 FIGURE 4.4 MEASURED LONGITUDINAL CRACKING (NON-WHEEL PATH) ............... 35 FIGURE 4.5 MEASURED VS. PREDICTED DISTRESSES (BASED ON THE DATA SUBMITTED BY THE ARA, ARA 2015B) ............................................................................... 36 FIGURE 4.6 PAVEMENT STRUCTURE ON I-95 SITE IN CHATHAM COUNTY ............... 37 FIGURE 4.7 PAVEMENT STRUCTURE ANALYSIS USING 1972 AASHTO INTERIM DESIGN GUIDE .......................................................................................................................... 38 FIGURE 4.8 PAVEMENT STRUCTURE ANALYSIS USING MEPDG .................................. 39
vi

FIGURE 4.9 REDESIGN OF I-95 SITE ...................................................................................... 40 FIGURE 4.10 AC THICKNESS (1972 DESIGN VS. MEPDG) ................................................. 41 FIGURE 4.11 E* FOR SMA AND SUPERPAVE....................................................................... 44
vii

EXECUTIVE SUMMARY
The Georgia Department of Transportation (GDOT) is in the process of evaluating the use of the AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG) for designing its new and rehabilitated pavement structures. GDOT has undertaken projects to establish the groundwork for the use of the MEPDG, including characterizing material properties, analyzing traffic loading, and calibrating the MEPDG performance prediction models for Georgia's local conditions and materials. A Georgia Long-term Pavement Performance (GALTPP) program was initiated by GDOT to provide a sufficient number of sites for the initial MEPDG local calibration, and, more importantly, to conduct long-term performance monitoring on the sites of GDOT's interest to support the performance evaluation and/or future MEPDG recalibration. The outcomes/findings will improve GDOT's practices of pavement design, material, construction, and maintenance. Currently, the GALTPP comprises 38 flexible pavement sites (17 LTPP and 21 non-LTPP sites) and 23 rigid pavement sites (11 LTPP and 12 non-LTPP sites). Various field and laboratory tests, including condition surveys in accordance with LTPP Distress Identification Manual, Falling Weight Deflectometer (FWD), Dynamic Cone Penetration (DCP) tests of the base and subgrade, bulk specific gravity measured on each layer, etc., were conducted on the non-LTPP sites, and documents (e.g., as-built plans and construction files) were gathered to provide the data needed for the calibration. These data are essential for further recalibration of the MEPDG. Therefore, it is essential to manage and maintain the data collected for the GALTPP program, including the existing data and the data to be collected in the future. In addition, the potential for characterizing non-standard designs and materials (e.g., micro-milling and Stone Matrix Asphalt (SMA)) used in Georgia must be identified using the MEPDG to provide suggestions
viii

on the implementation and future calibration.
This project consists of three consecutive one-year phases with each phase focusing on 1) a component for maintaining the GALTPP data and 2) the use of MEPDG for a specific design or material identified by GDOT. Phase 1 of this project focused on developing a database for the flexible pavement sites and evaluating the use of the MEPDG for designing Georgia's interstate pavement structures because they account for a major part of total capital investments on the roadways. Phase 2 will focus on extending the database to rigid pavement sites, and the potential topics for studying include warm mix asphalt, jointed plain concrete pavement (JPCP) design, etc., and will be further discussed with and confirmed by GDOT. Phase 3 will focus on the procedures for incorporating additional data (e.g., performance data) and sites (e.g., warm mix asphalt sites). The potential topics for Phase 3 will be determined at the end of Phase 2. The following are the major findings from Phase 1 of the project:
1) The data collected on GALTPP sites, including FWD DCP, etc., were gathered and carefully reviewed. The site locations were verified by comparing them (x-y coordinates included in GDOT's calibration study) to the core locations collected using GDOT's PDAbased core data collection application to ensure the sites can be correctly located for longterm performance monitoring. It was found that the site location data does not match the core location data. Thus, the site location data was corrected based on the first core located along the travel direction. The location data was further processed to obtain additional location information (e.g., RCLINK and milepoint) using GDOT's location reference system.
2) A database (GALTPP database) with location reference information was designed to store and manage the input parameters used in the MEPDG calibration, the condition survey data, ix

the testing data, and the documents collected on the GALTPP sites. GALTPP database tables and fields for flexible pavement were designed based on a relational database concept with geospatial information so it can be integrated into a GIS (Geographic Information Systems) platform. The data were processed and populated into the GALTPP database. In addition, a GIS project, along with add-in tools, was developed using the GALTPP database for visualizing the sites. 3) A review of GDOT's pavement condition survey data shows raveling is the predominate distress on Georgia's interstate pavements; in FY 2015, 41% of interstate segments were reported with raveling. Raveling is also an important performance indicator that triggers the need for maintenance on the porous friction course (e.g., Open Graded Friction Course (OGFC) or Porous European Mix (PEM)) on the surface layer, but it is not modeled in the MEPDG. 4) A total of 38 sites (17 LTPP and 21 non-LTPP sites) were used to calibrate the coefficients in the MEPDG transfer functions to eliminate bias and improve accuracy (i.e., reducing the standard error). Among them, five sites are on interstates. Compared to the other sites, these interstate sites exhibited low fatigue cracking (less than 3%) and moderate rutting (between 0.15 in. and 0.3 in.) at the end of pavement service interval (i.e., before pavement rehabilitation). The only site that exhibited more cracking (approximately 10% in 17 years) is on I-520, which has 7 in. of asphalt concrete layers. Based on the limited data, the measured distresses were within the distresses predicted at 50% reliability using the MEPDG.
5) A case study was conducted on I-95 in Chatham County based on the existing pavement structure. Using the MEPDG, the predicted distresses would reach the distress performance x

criteria (0.35 in. of rutting and 10% of fatigue cracking) at 95% reliability in 20 years. However, the observed distresses (0.25 in. of rutting and 3% of fatigue cracking) were close to the distresses predicted at 50% reliability. 6) Compared to the MEPDG, the current design procedure (1972 AASHTO Interim Design Guide; for brevity, called the 1972 Design Guide hereafter) is on the conservative side. The interstate pavement structure on the I-95 site was 10.17% under-designed when it was validated using the 1972 Design Guide. According to the 1972 Design Guide, to carry the 16.2 million heavy trucks, an additional 2 in. of asphalt base was needed. However, the design without the 2 in. of asphalt base passed all the performance criteria when it was validated using the MEPDG. 7) Though it is on the conservative side, the current 1972 Design Guide allows GDOT to replace only the top porous friction course in 10 to 12 years, and both the porous friction course and SMA layer in 20 to 24 years without the need to replace the underlying layers because the underlying layers are still structurally sound with few limited distresses. Analyses on multiple pavement service intervals (e.g., more than 20 years) based on GDOT's maintenance practices would help to determine the most cost-effective pavement structures. 8) Based on the field observation, SMA has better performance in terms of fatigue life on heavily traveled roads compared to Superpave. This benefit is not modeled in the current design procedure (1972 Design Guide) because both SMA and Superpave have the same structure coefficient of 0.44. This issue remains the same in the MEPDG. Moreover, using the Witczak predictive model in the MEPDG, the SMA has a slightly lower dynamic modulus than Superpave. This leads to higher predicted rutting and fatigue cracking,
xi

although the differences are small. However, this is contrary to the observed field performance in Georgia with better rutting and fatigue resistance of SMA. To move forward in maintaining an active GALTPP program, the following are recommended: 1) Because the pavements continue deteriorating, it is recommended that distress and FWD data be collected on an annual or biennial basis on the GALTPP sites to establish a long-term performance monitoring. Especially, it is recommended that cracking data be collected before and after resurfacing on the I-95 site in Chatham County, which will be resurfaced next year. Such data allows GDOT to validate the performance of this site and assess the development of cracking on micro-milled surface. 2) The 3D laser technology (e.g., 3D pavement data, video log images, etc.) can be used for collecting consistent and detailed pavement distress data on the GALTPP sites. The highresolution 3D laser data can be used for detecting cracks and quantitatively and objectively measuring raveling on the porous friction course surface. In addition, it can collect the cracking data before and after the micro-milling is performed on the porous friction course. These data are invaluable for assessing the development of top-down and bottom-up cracking on Georgia's pavements. 3) A raveling prediction model (including a measure for quantifying raveling) should be developed and incorporated into the life-cycle analysis of the interstate pavement design (for new and rehabilitated pavement structures). This will allow GDOT to reliably quantify raveling and identify the timing for adequate treatment(s), which is difficult using current visual inspection.
xii

4) It is recommended life-cycle cost analysis be performed based on GDOT's maintenance practices to determine the pavement structure design that is most cost-effective for the full life cycle.
xiii

ACKNOWLEDGMENTS The authors would like to thank the Georgia Department of Transportation (GDOT) for its support. The work conducted in this report was sponsored by the GDOT's Office of Research (Research Project 14-25). The authors would like to thank the support provided by the Office of Research and the Office of Materials and Testing. The strong support and valuable inputs provided by Mr. Moussa Issa, Mr. Steve Panaho, and Mr. Monzy Mathews of GDOT, and Dr. James Lai in the course of this research project are highly appreciated.
xiv

xv

1. INTRODUCTION
1. 1 Background and Research Need The Georgia Department of Transportation (GDOT) is in the process of evaluating the use of the AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG), developed under the National Cooperative Highway Research Program (NCHRP) Project 1-37A (NCHRP 2000), for designing its new and rehabilitated pavement structures. The MEPDG models pavement responses (stresses, strains, and deflections) using traffic loading, material properties, and environmental data, and it relates the cumulative damage to field-observed pavement performance empirically using pavement distress transfer functions (or distress prediction models). GDOT has undertaken several projects to establish the inputs for the MEPDG, including conducting tests to characterize material properties, studying traffic load spectra, etc., and, also, conducted verification and local calibration of the MEPDG performance models for use in Georgia (ARA 2015a). During the verification, it was found that the number of LTPP sites in Georgia is insufficient to cover the range of pavement structures, materials, and other design features commonly used by GDOT, and the levels of distress exhibited on these LTPP sites are inadequate for the calibration of the MEPDG. Therefore, GDOT initiated a Georgia Long-Term Pavement Performance (GALTPP) program, which includes LTPP sites in Georgia and additional sites (referred as non-LTPP sites) to cover common design features used in Georgia for support of the MEPDG calibration (ARA 2015a). Currently, the GALTPP program comprises 38 flexible pavement sites (17 LTPP and 21 non-LTPP sites) and 23 rigid pavement sites (11 LTPP and 12 non-LTPP sites). Extensive field and laboratory testing, including condition surveys in accordance with LTPP Distress Identification Manual, Falling Weight Deflectometer (FWD), Dynamic Cone Penetration (DCP) tests of the base and subgrade, bulk specific gravity
1

measured on each layer, etc., were conducted on the non-LTPP sites to obtain input data, including pavement design, material properties, performance data, etc., for the MEPDG.
Though the initial calibration was completed, it is recognized that the recalibration of the MEPDG is still needed in the future as MEPDG performance models (e.g., reflective cracking model) are improved, as more distress data becomes available over time, and as new pavement methods and materials are implemented in Georgia. The rich data collected on the GALTPP sites (both LTPP and non-LTPP sites) are valuable to GDOT and essential for support of MEPDG recalibration in the future. Besides the current sites, the GALTPP program is expected to include additional sites in the future for evaluating the effects of different designs, materials, construction methods, maintenance levels, etc., on pavement performance. For example, GDOT has built research sites with new methods and materials, such as the use of micro-milling, warm mix asphalt (WMA), and crumb rubber modified asphalt. These research sites should be documented, monitored, and tracked through the GALTPP program. In addition, studies are required to evaluate the feasibility of modeling non-standard methods and materials used in Georgia using the MEPDG. Therefore, the objectives of this project are 1) to maintain the data (e.g., LTPP survey and FWD) that has been collected and will be collected on the GALTPP sites to support the recalibration, 2) to include pavement research sites built with new designs, materials, etc., into the GALPP program to document and monitor their long-term performance, and 3) to identify the potential for the characterization of non-standard methods and materials using the MEPDG to provide suggestions on the MEPDG implementation and future recalibration.
2

This project consists of three consecutive one-year phases with each phase focusing on one component for maintaining the data collected for the GALTPP program and one specific method and material identified by GDOT. Table 1.1 lists the work by phases. This allows GDOT to prioritize the methods and materials to study in this project and provides the flexibility to study the sites that are relatively new in later phases. Phase 1 of this project focused on developing a GALTPP database for maintaining the data collected on the flexible pavement sites and evaluating the design of pavement structure on Georgia's interstate highways using the MEPDG. Phase 2 will focus on extending the database onto rigid pavement sites, and the potential topics for studying, such as jointed plain concrete pavement (JPCP), will be further discussed with and confirmed by GDOT. Phase 3 will focus on the procedures for incorporating additional data (e.g., performance data) and sites (e.g., WMA sites). The potential topics for Phase 3 will be determined at the end of Phase 2.

Table 1.1 Work by Phases

Maintaining GALTPP data

Potential Topics

Phase 1 Flexible pavement sites

Interstate highway

Phase 2 Rigid pavement sites

To be determined (e.g., jointed plain concrete pavement)

Phase 3 Incorporating research sites

To be determined

1.2 Significance of Research Maintaining the data collected for the GALTPP program will allow GDOT to track and share data collected on the sites with different designs, materials, construction methods, and maintenance levels to support quantitative assessment of their effects on long-term pavement
3

performance. The GALTPP database will serve as one of the most important sources of data for further validation and calibration of the MEPDG models and the evaluation of the effects of different pavement designs, materials, etc. The outcomes/findings can be used to improve GDOT's practices for pavement design, material selection, construction methods, and maintenance strategies. In addition, the outcomes on the potential for characterizing nonstandard methods and materials used in Georgia using the MEPDG will enable GDOT to better utilize the MEPDG for understanding the distresses based on different designs and materials.
1.3 Research Objectives and Scope The objectives of Phase 1 of this project were to develop a GALTPP database for maintaining the data collected on the flexible pavement sites and 2) evaluate the design of interstate pavement structure using the MEPDG. The specific activities for each work task are presented below: 1) Work Task 1: Manage the data collected on the flexible pavement sites.
In this task, the Georgia Tech research team worked with GDOT's GALTPP Task Force and the Office of Research to gather the data (including LTPP distress survey data, FWD, DCP, and coring data) that were used to support the calibration of the MEPDG. A relational database (GALTPP database) with location references was designed and developed for more efficient and easier data management and manipulation; the data for the flexible pavement sites was carefully reviewed and populated into the developed GALTPP database. In addition, a GIS project was developed to visualize the GALTPP sites (including LTPP and non-LTPP sites) and the data from various sources.
4

2) Work Task 2: Identify pavement structure designs to be evaluated using the MEPDG and gather data for the sites. Because interstate highways account for a major part of capital investment, GDOT's GALTPP Task Force set the focus of this phase to evaluate the feasibility of designing interstate pavement structures using the MEPDG based on the procedure and input parameters recommended in the Georgia ME Design User Guide (ARA 2015b). In addition, the distresses on the interstate highways were studied based on GDOT's pavement condition survey conducted in FY 2015 to better understand the distresses on the interstate highways.
3) Work Task 3: Evaluate the feasibility of designing Georgia's interstate pavement structures using the MEPDG. This work task is to evaluate the feasibility of designing Georgia's interstate pavement structures using the MEPDG. This includes verifying the distress predicted by the MEPDG and comparing the pavement structures designed by using the MEPDG and the 1972 AASHTO Interim Design Guide (for brevity, called 1972 Design Guide hereafter) (AASHTO 1972) to provide suggestions on the implementation. The major subtasks are listed as follows: o Review the distresses data used for calibrating the MEPDG performance models, especially the data on interstate sites; o Run the AASHTOWare Pavement ME Design to predict pavement performance (e.g., fatigue cracking and rutting) and compare the predicted and observed distresses on interstate sites; o Compare the pavement structures designed by the MEPDG and the 1972 Design Guide and provide suggestions on the implementation of the MEPDG.
5

4) Work Task 4: Prepare final report. This task is to summarize the findings of Phase 1 and make recommendations.
1.4 Organization of this Report This report is organized as follows: 1) Chapter 1 introduces the background, significance, scope, objective, and work tasks of this
project. 2) Chapter 2 presents the development of a GALTPP database for a) the data, such as distresses,
coring, etc., collected on the GLAPP sites, b) a data repository, including a database design and structured folder, for the collected data and the inputs to the MEPDG, and c) a GIS project for integrating and visualizing the data. 3) Chapter 3 describes current pavement structure design and maintenance practices on the interstates and analyzes the distresses on interstates based on pavement condition evaluation data collected by GDOT in FY 2015. 4) Chapter 4 evaluates the design of Georgia's interstate pavement structures, which is comprised of porous friction layer (e.g., Open Graded Friction Course (OGFC) or Porous European Mix (PEM)) and Stone Matrix Asphalt (SMA), using the MEPDG. The differences between the MEPDG and the 1972 Design Guide are discussed. 5) Chapter 5 summarizes the findings of this project and makes recommendations.
6

2. MANAGEMENT OF GALTPP DATA
The goal of the GALTPP program is to provide data that supports quantitative evaluation of the effects of various pavement designs, materials, and maintenance strategies on pavement performance. The findings can help improve GDOT's practices for pavement design, material selection, construction methods, and maintenance strategies that will lead to more cost-effective and better performing pavements. To achieve this goal, the GALTPP program needs to gather, process, and share the data that describes the pavement structures, material properties, traffic loads, long-term performance, etc., on each site. This chapter presents the data gathered for the GALTPP program, suggestions on the location references, the database designed to store the data gathered for GALTPP program, and a GIS project developed to facilitate the data visualization and integration.
2.1 Data Gathered for the GALTPP Program This section describes the data gathered for the GALTPP program with a focus on flexible pavement sites, including 17 LTPP and 21 non-LTPP sites. For the LTPP sites, the data included in the LTPP database were used in support of the MEPDG calibration. For the non-LTPP sites, project construction files and GDOT's Pavement Condition Evaluation Systems (PACES) (GDOT 1993) data were gathered for each site. In addition, field data collection and laboratory testing was performed in 2014 by the Applied Research Associates (ARA) and the National Center for Asphalt Technology (NCAT) to collect the inputs needed for the MEPDG calibration (ARA, 2015a). The collected data included (a) condition surveys in accordance with the LTPP Distress Identification Manual (FHWA 2003), (b) FWD deflection basin testing, (c) DCP tests of
7

the base and subgrade, and (d) drilling cores. The appropriate input parameters were then determined for each of the LTPP and non-LTPP sites based on the field laboratory testing data. A final set of MEPDG input parameters and field and laboratory testing data were received in April, 2016. The following describes the data included in the data set.
Distress surveys were conducted in accordance with the LTPP Distress Identification Manual (FHWA 2003) to identify the severity level and extent of distresses observed on each sites. It is noted that there were typically more than one LTPP survey on each LTPP site. However, they were conducted in earlier years when the distresses were limited. The distress data were carefully reviewed, and the data with an irregular trend were removed from the calibration. For non-LTPP sites, there was only one survey conducted in 2014. However, the amount of cracking on non-LTPP sites was much higher than the one on LTPP sites. In addition, some PACES data on non-LTPP sites were converted into the distresses defined in the LTPP (FHWA 2003) and used in the calibration.
FWD deflection basin measurements were made every 50-feet in the outside wheel path for each non-LTPP site; the data was stored in a proprietary format (.F25)
A total of nine cores were taken on each non-LTPP site. Three cores were taken in distressed areas directly over the cracks to determine whether the cracks initiated from the top or bottom of the HMA layers. The other six cores were taken randomly in areas without distresses. Individual layer thicknesses were measured from the cores in the lab. The bulk specific gravity of each layer was measured in accordance with AASHTO T 166, "Bulk Specific Gravity (Gmb) of Compacted Hot Mix Asphalt (HMA) Using Saturated Surface-Dry Specimens." The maximum specific gravity of each layer was measured in accordance with AASHTO T 209: "Theoretical Maximum Specific Gravity
8

(Gmm) and Density of Hot Mix Asphalt (HMA)." The asphalt concrete content was

measured for selected layers near the bottom of the pavement. The bulk specific gravity

and maximum specific gravity of each layer were listed in an Excel sheet.

DCP measurements were performed and recorded at three core holes on each non-LTPP

site. The DCP penetration rates (mm/blow) were then used to estimate the in-place

resilient modulus for the unbound layers using Equation (1), which was developed by

ARA (ARA, 2015a). The raw data such as penetration and calculated resilient modulus

were stored in an Excel sheet.

=

0.145



17.6



(

292



0.64
) 1.12

(1)

Pictures taken at the non-LTPP sites and cores were organized by site and included in the

data set. It also included MEPDG files for all the GALTPP sites.

It is noted that these various field and laboratory testing data need to be integrated by site location). Thus, it is crucial to accurately record the site or location information. It is
that GDOT's PDA-based core data collection application (Wang and Tsai 2013) be used for collecting coring data and photos on each site. In this way, coring location and photos were
automatically tagged with x-y coordinates from the high-accuracy, built-in GPS. For the data, using a template, as shown in
Table 2.1, is suggested for recording the data with consistent location information for easy data integration. Columns 1-9 show the location information and Column 10 shows the data if collected in a proprietary format. This is especially important when the data is collected by different crews (or contractors) at different times.
9

Table 2.1 Example Template for Recording Data with Location Information

(1)

(2) (3) (4)

(5)

(6)

(7)

(8) (9)

(10)

Site ID X Y County Route Route Direction Lane Milepoint File

No

Alias

No

R1

Chatham 0405

Pos

2

8.1

.\\

R2

Camden 0405

Neg

2

20.1

.\\

2.2 Correction of Site Locations The site locations (x-y coordinates) were verified to ensure the sites can be located correctly for conducting long-term performance monitoring. This is especially important for the non-LTPP sites that do not have any signage to indicate site locations. The site locations were verified by comparing them to the core locations recorded using GDOT's core data collection application that runs on a GPS-enabled PDA (Wang & Tsai 2013). Figure 2.1 shows all the non-LTPP sites for flexible pavements and the core locations. It was found that the site locations don't match the core locations; that is, the sites were outside the area where the cores were taken for the specific site. It was determined that the core locations recorded by GPS are more accurate than the site locations included in a previous study (ARA 2015a). For example, the flexible pavement site on I-85 was found wrongly located near Milepost 66 on concrete pavement based on the site location data; however, the cores were correctly located near Milepost 68 on asphalt pavement using the location data recorded by GPS, as shown in Figure 2.2. Therefore, the site locations were corrected using the first core located in the travel direction. In addition, the locations were processed to obtain the route and milepoint information using GDOT's linear reference system (LRS). The corrected site location information is listed in Appendix I. It is noted that the
10

research team obtained core location information for twenty sites. Core information for the remaining sites will be requested and obtained to verify the locations for these sites.
Figure 2.1 Site and Core Locations

(a) Site location

(b) Core location(s)

Figure 2.2 Verification of Site Location (I-85 site)

11

2.3 Design of GALTPP Database A database is needed to store and organize various data collected for the GALTPP program and to manage the data efficiently. This database will serve as a centralized source of the GALTPP data to be used for studying the effects of different pavement designs, materials, etc., on pavement performance. To take advantage of information technologies, a relational database (GALTPP database) with location reference information was designed to house the MEPDG input parameters used in the calibration, the field and laboratory testing data collected on the GALTPP sites, and other information gathered on the sites. With the location reference information, the GALTPP database can be readily tied into a GIS platform for visualizing the data on a map and performing spatial query. The current GALTPP database was developed using the geopersonal database in ArcGIS for its easy access and GIS integration functions. The database will be enhanced throughout the course of this project based on GDOT's comments; the final version can be created in an enterprise database (e.g., Oracle database or SQL server) based on GDOT's requirements. The design of the GALTPP database involved identifying data elements to be stored, designing a database architecture that relates foreign and primary keys and table structures, and designing location reference information to be used in the GALPP database. Figure 2.3 illustrates the high-level conceptual design of the GALTPP database. The GALTPP database is the central place for storing data on both LTPP sites and non-LTPP sites. Although the distress data collected based on both LTPP and PACES protocols can be available on a site, the performance data stored the GALTPP database followed the LTPP format. A conversion between LTPP and PACES distresses was recommended in Georgia's MEPDG User Guide (ARA, 2015b). In addition, the GALTPP database is not intended to duplicate the completed LTPP database or the existing GDOT PACES database; instead, it was designed for easy access and
12

management of the MEPDG input parameters used in the calibration and the field and laboratory testing data collected on the GALTPP sites. The data to be stored in the GALTPP database can be primarily organized into eight different categories; (1) site information, (2) pavement structures, (3) pavement performance, (4) traffic, (5) material properties, (6) environment (or climate), (7) testing data, and (8) files to cover various data collected on the GALTPP sites. The data to be stored in each category were briefly described in this section. Appendix II provides a tabular listing of the tables in the GALTPP database.
Figure 2.3 Conceptual Design for GALTPP Database 13

Site information The site information table contains the location reference information. The table contains both x-y coordinates (Columns 2 and 3 in Table 2.1) and GDOT's Road Characteristic Link (RCLINK) (Columns 4 to 9 in Table 2.1) for identifying site location. RCLINK (defined using county, route type, route number, route suffix) is a unique identifier for a route. Each RCLINK typically begins with zero in its linear measure (e.g., milepoint). This means for the same route the milepoint is reset to zero at the county boundary. RCLINK along with milepoint(s) can be used to identify the location of a point or linear event on a route. In addition to location information, data such as functional class and number of lanes are also stored in the table. A unique identifier (e.g., site_ID) was assigned to each individual GALTPP site. For LTPP sites, the SHRP_ID is used; for nonLTPP sites, a unique ID is assigned.
Pavement structures Pavement structure data, including total number of layers, layer number, layer type, and layer thickness, were stored in a table. It is noted that for calibration purposes, a site can be used as a new pavement structure at the beginning and later as a rehabilitated pavement structure after a treatment is applied on it. Any construction-related change to the pavement structures is recorded by defining a new identifier for the construction (e.g., construction ID) in the table. Thus, site_ID alone cannot uniquely identify a pavement structure on a site. Instead, the combination of site_ID and construction_ID must be used to identify a pavement structure on a site.
14

Pavement performance Distress data is stored in a format that is similar to the one in the LTPP database. It is noted that not all distress data can be used for a MEPDG calibration. The distress data with irregular trends were considered as outliers, and were excluded when calibrating any MEPDG prediction model. For example, the decrease of distresses without proper treatment was considered unreliable and was not used in the calibration. Therefore, a field is designed in the table to indicate whether or not the observed distresses can be used in a calibration.
Traffic Tables were designed to store traffic data on each site, including AADTT, growth rate, vehicle class distribution, etc.
Material properties Tables were designed to store the material properties for each layer, including aggregate gradation for asphalt mix, effective binder content, air voids (at time of construction), total unit weight, asphalt binder data, etc. These tables typically include a layer number to identify the material properties for each layer.
Environment (or climate) A table was designed to store environmental/climatic factors used in the MEPDG, including elevation, weather station, and groundwater depth.
Testing data Field and laboratory testing data, such as DCP, bulk specific gravity, and maximum specific gravity, were stored in different tables. It is noted that the data stored in these
15

tables are considered raw data and are not necessarily the same as the MEPDG input parameters. Files Tables in this category were designed to provide a link to the documentation (e.g., pdf file), images, and data collected in a proprietary format (e.g., FWD file). Instead of storing these files in the database, workspaces were designed to house electronic files, such as the as-built plans, the construction records, the FWD files, etc.; tables were designed to store the file path for retrieving the files. These tables typically contain site ID, date of data collected, file type, file path, and x-y coordinates (if available).
2.4 Visualization of GALTPP Data in a GIS Project With the location references, a GIS project was developed using ArcMap to allow the users 1) to visualize the geographic distribution of candidate sites, and 2) to integrate with the data from other sources. The use of GIS allows GDOT to facilitate the coordination among GDOT's offices. The functions in the GIS project are described in the cases below:
Case 1: Visualize GALTPP sites Using GDOT's LRS and the dynamic segmentation function in GIS, GALTPP sites were spatially integrated onto a map with other data, such as the pavement design data and soil data. GDOT's engineers can navigate the map to visualize information on the map, as shown in Figure 2.4. With their knowledge of Georgia's soil, weather, and pavement conditions, they can effectively identify any issue in the geographic distribution of the GALTPP sites. For example, the distribution of the sites in northern and southern Georgia may be a concern for the GALTPP sites because of the significant differences in 16

the geologic conditions. In addition, a cluster of the sites in certain areas (e.g., in one district) can be identified effectively using visualization. Case 2: Facilitate the communication among different offices Coordination among GDOT's offices is crucial for maintaining the GALTPP sites in the long-term. Especially, some of the sites will be resurfaced in the near future, and these activities should be coordinated between the Office of Research, the Office of Materials and Testing, and the Office of Maintenance. A pavement condition evaluation can be planned before maintenance and rehabilitation activity, and the maintenance record can be gathered and recorded in the GALTPP program. Currently, the Office of Maintenance can output the planned resurfacing projects in an Excel format using GDOT's PMS. This file can be sent to the Office of Research and with the location information (county, route number, route suffix, milepoint from, and milepoint to) the planned resurfacing projects can be mapped using the GIS project to identify the GALTPP sites that will be resurfaced. The Office of Research, the Office of Materials and Testing, and the Office of Maintenance can coordinate on the data to be collected on the GALTPP site(s) before being resurfaced and other activities.
17

Figure 2.4 An example of visualizing data in GIS 18

3. PAVEMENT DESIGN, MAINTENANCE PRACTICES, AND PERFORMANCE OF GEGORIA'S INSTERSTATE HIGHWAYS
Interstate highways account for a major part of the capital investment in the transportation infrastructure. The decisions on interstate pavement structure design, maintenance, and rehabilitation have a great financial impact. Thus, GDOT's GALTPP Task Force set the focus in the Phase 1 of this project to evaluate the feasibility of designing Georgia's interstate pavement structures using the MEPDG to provide inputs on the MEPDG implementation. This chapter briefly describes pavement structure design and maintenance practices currently used on Georgia's interstate highways. In addition, the predominant distresses on interstate highways were reviewed base on GDOT's pavement condition evaluation data collected in FY 2015 to identify the performance indicator(s) needed on interstate highways.
3.1 Pavement Design and Maintenance Practices Georgia's interstate highways have commonly been constructed with four different asphalt concrete layers, a 0.875 1.25-in.porous friction course; a 1.5-in. 12.5-mm SMA; a 2-in.19-mm binder layer, and a 4-10 in. 25-mm base layer on the top of graded aggregate base (GAB). Figure 3.1 shows the current typical pavement structure design for Georgia's interstate highways. The use of porous friction course and SMA is a somewhat unique design, and it provides safety during wet weather and durable pavements under heavy traffic volume.
GDOT's use of porous friction course (e.g., D mix) dates back to the 1970s. Porous asphalt layers have a high air void content (10-20%), which allows rapid removal of surface water in
19

light to moderate rain through the pores. Thus, it has been used to enhance safety on the roadways during wet weather, such as reducing splash and spray, improving visibility of traffic stripes, etc. Since the 1990s, GDOT started using OGFC and PEM, which include stabilizing fibers and polymer modified asphalts (PMA) and perform better than D mix. SMA is a gap graded mix with a high concentration of course aggregates that increase stone-to-stone contact, create a more efficient network for load distribution, and make it a good choice of mix for highvolume roadways. The stone-to-stone skeleton held together by rich asphalt cement is the key to its ability to withstand rutting and fatigue cracking. GDOT became interested in the use of SMA after the European Asphalt Study Tour of 1990 because of SMA's potential durability improvements (Watson et al. 1995); GDOT began researching the viability of using SMA in Georgia in 1990. Test sites were built on I-85 in 1991 to evaluate the performance of SMA; studies were also conducted to determine optimized SMA mix design (Watson et al. 1995; Barksdale 1995; Jared 1997a and 1997). Based on previous studies (Jared 1997a and 1997b), SMA is expected to have 30-40% less rutting, a 30-40% longer fatigue life, and a lower annualized cost than standard mixes. Since the 1990s, a 1.5-in of SMA has been used on interstates due to the high traffic volumes that interstates in Georgia carry.
Figure 3.1 Typical pavement design in Georgia's interstate highways
20

GDOT has had an active resurfacing program since the 1980s. The resurfacing program focuses on the use of thin-resurfacing (1.5 in.) to replace a worn-out surface layer at the right time to prolong pavement life. Based on GDOT's experiences, the porous friction course typically wears out after 10 to 12 years, while the underlying SMA layer is still sound. Prior to 2007, the common practice for replacing the worn-out open-graded layer was to mill and replace this layer together with the underlying 1.5 in of SMA using conventional milling (Lai et al., 2012; Tsai et al., 2014; Tsai et al., 2015). Milling and replacing only the porous friction course (0.875 1.25 in. of OGFC or PEM) using conventional milling was not considered because of the concern about placing such a thin, open-graded mix on the resulting surface texture from conventional milling. With conventional milling, the ridge-to-valley depth (RVD) caused by the milling teeth can be 0.25 in. or greater. Placing the open-graded mix on this irregular surface concerned GDOT about the improper channelization of water and insufficient bonding between the coarser aggregate used in the open-graded mix and the milled surface. The purpose of milling and replacement of a1.5-in. of sound SMA layer, in addition to the worn-out porous friction course, is to 1) provide good bonding between the open-graded surface layer and the rough milled surface and 2) reduce the potential for water entrapment in the valleys created by the milling head teeth in the rough milled surface texture. In 2007, GDOT developed a new, cost-effective method ( micro-milling and thin-overlay operation) to replace only the porous friction course directly over the micro-milled surface without removing the sound underlying layer (Lai et al., 2012; Tsai et al., 2015). With the new method, GDOT's maintenance practices have been changed to utilize the new method to replace only the worn-out OGFC or PEM when the underlying SMA is still sound. In addition, the 2 in. of SMA is expected to withstand the micromilling operation and provide sufficient structure capacity. As shown in Figure 3.2, the OGFC is
21

expected to be replaced in 10-12 years. With the use of micro-milling and thin-overlay, only the OGFC layer will be replaced directly on the top of SMA layer. As the OGFC deteriorates in another 10-12 years, both the OGFC and SMA layers will be replaced. This provides a costeffective approach to maintaining Georgia's interstates.

OGFC

Replace OGFC using Micro-milling

& thin-overlay

Replace OGFC and SMA

SMA

Time

10

20

Figure 3.2 Interstate life-cycle maintenance activities

3.2 Predominant Distresses on Interstates Since the 1980s, GDOT has been conducting an annual pavement condition evaluation of its 18,000-centerline miles of roadway based on its PACES. This section presents the predominant distresses recorded on interstate highways based on the PACES data collected in FY 2015. PACES surveys involve recording the severity and extent of various types of pavement surface distresses. They include rutting, load cracking, block cracking, reflective cracking, raveling, edge distress, bleeding/flushing, corrugation/pushing, loss of site, and potholes/patches/localized failure, as listed in Table 3.1 Distresses in PACESTable 3.1. It is noted that a walking survey is conducted for cracking; the survey is of a 100-foot representative sample location per mile-long segment. The distresses are recorded for each segment (which is about one mile long), then
22

aggregated/averaged to obtain the representative pavement condition for a project (typically several miles long).

Table 3.1 Distresses in PACES

Distress

Unit

Severity

Load Cracking Block Cracking
Reflection Cracking Edge Distress Rutting Patches/Potholes/Local failure Bleeding Raveling Corrugation Loss of Site

% % Number of cracks Length in foot % 1/8 inch
Number % % % %

1, 2, 3, 4 1, 2, 3 1, 2, 3 1, 2, 3
1, 2, 3 1, 2, 3 1, 2, 3 1, 2, 3

Sample Location
100-ft 100-ft
100-ft 1-mie 100-ft
1-mile 1-mile 1-mile 1-mile 1-mile

The distresses recorded at the segment level for the interstate highways in Georgia in FY 2015 were analyzed. The pavement distress evaluation is performed according to the GDOT's PACE protocol (GDOT 1993). There were a total of 1,495 interstate segments surveyed in FY 2015. Figure 3.3 shows the rating distribution for these segments. Of the segments, 88% of them had a rating greater than or equal to 70, and 12% of them had a rating less than 70. 3% of the segments had a rating less than 60.

23

3% 9%
16% 53%
19%

90-100 80-89 70-79 60-69 0-59

Figure 3.3 Interstate rating distribution (FY 2015)
Figure 3.4 shows the percent of segments for each type of distresses. Different colors are also used to highlight the percent of segments with different severity levels. It can be seen that rutting was the most commonly reported distress (46%) on interstates; however, it is not a concern from pavement maintenance point of view because the majority of the distresses were 1/8 in. (highlighted in green). There were 17% of segments with in of rutting (highlighted in blue), and only 1% of the segments had 3/8 in. of rutting (highlighted in red). Raveling was the second most frequently reported distress (41%) on interstates because of the use of porous friction course. It is also an important performance indicator for triggering maintenance needs because severe raveling raises safety concerns. A rapid deterioration of raveling in terms of its severity and extent was observed in the PACES data. In the current visual inspection, it was difficult to capture the early-stage raveling. Inconsistencies in the severity level and extent have been observed in the data. For example, some segments had Severity Level 2 raveling in one year, but they became severity level 1 in the following year. This makes it difficult to reliably determine the timing for preservation treatments (e.g., fog seal). In addition, raveling is currently not modeled in the MEPDG. Approximately 35% of the segments were reported with load cracking. The extents ranged from 2% to 100% with an average of 23%. It is noted that load cracking is
24

defined as cracks in wheel paths within the 100-ft sample location. Studies (Dauzats & Rampal, 1987; Craus et al., 1994; Uhlmeyer et al., 2000) show that wheel path longitudinal and fatigue cracks in the thicker asphalt concrete pavements more often initiate from the top of the wearing course downward. Based on GDOT's experience, there has been limited bottom-up cracking observed on interstate highways. It is noted that cracks initiated with truck rim cut or at the opengraded course are also considered as load cracking, as shown in Figure 3.5. The scratches on the surface can develop into wide cracks because of the loose aggregates. Thus, the load cracking reported in PACES does not necessarily relate to the bottom-up cracking. They could be developed on the open-graded course only or into the SMA layer.
Figure 3.4 Interstate distresses frequency (FY 2015)
25

Figure 3.5 Photo taken on I-95 Northbound MP 13-14 Chatham County in 2014 Current 3D sensing technology has the unique capability to capture accurate raveling data on the surface of porous friction course. These data are invaluable for developing the raveling prediction model for Georgia's interstates. In addition, the technology can be used to detect cracking with location reference at different stages (prior to the micro-milling, after the micromilling, and after resurfacing is performed) for assessing the development of top-down and bottom-up cracking. These data are also invaluable for studying the impact of the cracks on milled surfaces and the resurfacing performance, and for evaluating the effectiveness of the crack pre-treatments (such as cracking sealing). Figure 3.6 illustrates the use of 3D sensing technology for collecting and tracking (a) cracks on a raveled surface before micro-milling; (b) cracks on the micro-milled surface; and (c) the performance on the resurfaced surface.
26

(a) Before milling

(b) After milling

(c) After resurfacing

Figure 3.6 Photos taken on I-95 before and after resurfacing

In summary, raveling is the predominate distress on Georgia's interstates with porous friction course, but the current MEPDG is not capable of predicting raveling. Therefore, developing a raveling prediction model and incorporating it into the life-cycle analysis of the interstate pavement structure design (for new and rehabilitated pavements) is important for GDOT. In addition, the current visual inspection cannot reliably quantify raveling to identify the timing for adequate treatment(s). Current 3D laser technology with high-resolution data covering a full lane-width has the unique capability to quantitatively and objectively measure raveling on the surface of porous friction course. In the meantime, it can also collect cracking data before and after micro-milling is performed on the porous friction course. These data are invaluable for developing a distress prediction models for this particular pavement type. Furthermore, it can potentially be used for assessing the development of top-down and bottom-up cracking on Georgia's interstates.

27

28

4. EVALUATION OF DESIGN OF GEORGIA'S INTERSTATE PAVEMENT STRUCTURES USING DIFFERENT METHODS
Currently, GDOT designs its new and rehabilitated pavement structures in accordance with the 1972 Design Guide. It is an empirical method based on the AASHO Road Test equations that relate the loss in pavement serviceability to the pavement structures and load applications. While the 1972 Design Guide has been successful implemented for designing Georgia's pavement structures, it is recognized that there exist some limitations, including limited traffic inputs, a limited number of pavement test sites, a limited set of materials, and one climatic condition. In addition, it is difficult to relate the design to its performance (e.g., surface distresses). On the other hand, the MEPDG, developed under the NCHRP Project 1-37A (NCHRP 2000), is considered a more advanced pavement design tool. With its basis in empirical performance calibrations and mechanistic principles, resulting designs are considered to produce improved thickness estimates over the traditional empirical designs. GDOT has calibrated the MEPDG performance prediction models to Georgia's conditions and materials for flexible and rigid pavements using the sites in the GALTPP program. The calibration followed the procedure in the AASHTO MEPDG Local Calibration Guide (NCHRP 2004) to determine Georgia's calibration coefficients for eliminating the bias in using the global coefficients and to improve the accuracy (i.e., reducing the standard error). A total of 38 sites (17 LTPP and 21 non-LTPP sites) were used for calibrating the flexible pavements. The standard errors of the estimate for fatigue cracking and rutting are 5.8% and 0.105 in., respectively. These values are comparable with the ones reported in the global calibration and suggested in the local calibration guide (NCHRP 2004).
29

This chapter first reviews the distresses observed on interstate sites and compares them to the distresses predicted by the MEPDG to verify the accuracy of using the MEPDG performance prediction models on interstate sites. Second, a case study was conducted on a site on I-95 in Chatham County to analyze the pavement structures using different methods (1972 Design Guide and the MEPDG) to provide insight into the differences and their implications. Finally, the characterization of SMA in the MEPDG is discussed.

Table 4.1 Georgia's Calibration Coefficients

Transfer

Function

Coefficient

K1

AC Rutting

K2

K3

Coarse-Grained,

Subgrade Rutting

Bs1 Fine-Grained,

Bs1

AC Fatigue Cracking

K1 K2 K3

C1

Bottom-up Cracking

C2

C3

C1

Top-down Cracking

C2

C3

Thermal Cracking

Bt1 Bt3

1. Unmodified HMA mixtures

2. Polymer Modified Asphalt mixtures

3. Use global values

Global Value
-3.35412 1.5606 0.4791
1.0
1.0
0.007566 3.9492 1.281
1.0 1.0 6,000 7 3.5 0 1.5 1.5

GDOT Value

Neat1

PMA2

Mixtures

Mixtures

-2.45

-2.55

1.56063

1.56063

0.30

0.30

0.50

0.30

0.000653

0.00151

3.94923

1.2813

2.2

2.2 6,0003
73 3.53 03

35

45

35

45

4.1 Review of MEPDG Predicted Distresses on Interstate Sites A total of 38 sites were used for calibrating Georgia's transfer coefficients for flexible pavements. Among the 38 sites, five sites were on interstate highways. Figure 4.1 shows these five sites: three sites on I-95, one site on I-85, and one site on I-520. These five sites include both new
30

pavement design and overlay design. It is noted that these sites were built prior to 1996 (three sites were built in the 1970s and two sites in the early 1990s); SMA mixtures were not used on these sites. Therefore, they may not represent the actual performance of interstate pavements built with the recent pavement materials (e.g., SMA). This means additional interstate sites can be included to monitor the performance of SMA on Georgia interstate highways and to be considered in the calibration of coefficients for PMA mixtures.. Table 4.2 lists the pavement design of these five sites. It is noted that when the overlay design is modeled in the MEPDG, a 0% fatigue cracking was assumed on the existing milled pavement surface. This implies the underlying layers are sound without bottom-up or top-down cracking.
Figure 4.1 GALTPP sites 31

Site ID Route County Construction Year Pavement Design
Overlay Year Overlay Design

4112 I-95 Camden 1977

Table 4.2 Interstate Sites

4113

F9

I-95

I-85

Camden

Fulton

1977

-

F12 I-95 Chatham 1994

F19 I-520 Richmond 1995

Dense Graded HMA (3.1") Dense Graded HMA Base (12.7")
Poorly Graded Sand; A-3

Dense Graded HMA (3.6") Asphalt Stabilized/Treated Base (11.5") Poorly Graded Sand with Silt; A-3

1998

1998

Dense Graded HMA Dense Graded HMA

(1.8")

(1.8")

-

Porous friction layer Dense Graded

(0.75")

HMA, Type E (1.5")

Dense Graded

Dense Graded

HMA, Type D (1.5") HMA, Type B (2")

2002
Open Graded Friction Course (1.7") Dense Graded SMA with RAP, PG 76-22 (1.4") Milling to remove existing HMA surface Dense Graded HMA (1.4") Dense Graded HMA (1.6") Dense Graded HMA Base (8") Granular Aggregate Base (12") Clayey Silt

Dense Graded HMA Base (6") Granular Aggregate Base (14") Silty and Sandy Clay, A-2-4
2002
Porous friction layer
Dense Graded HMA, SMA (1.5")
Dense Graded HMA, Type D (2.25")

Dense Graded HMA, Type C (4") Granular Aggregate Base (12")
-

32

4.1.1 Observed Distresses This section presents the observed distresses on the interstate sites based on the data used to calibrate Georgia's coefficients (ARA 2015b). Figure 4.2 shows the measured fatigue cracking on all of the 38 flexible pavement sites with the interstate sites highlighted in red. It is noted that the magnitudes of the cracking occurring on the interstate and non-interstate sites are different. Compared to the non-interstate sites, the interstate sites exhibited less fatigue cracking. Less than 3% of fatigue cracking were recorded on the interstate sites, except for the site on I-520, which has less than 10 in. of AC. It is noted that other than the LTPP sites, most of the nonLTPP sites were resurfaced approximately every 11.6 years (Tsai, 2015). Thus, there were fewer cracks recorded with an age greater than 11.6 years. The resurfacing would remove distresses (e.g., cracking and rutting) on the surface layer, which makes it difficult to accumulate cracking data. It is noted that there were cracks recorded with an age greater than 15 years. This means these sites had not been resurfaced in more than 15 years, which is much longer than GDOT's average resurfacing years (Tsai, 2015). In addition, some interstate sites show almost no cracking after more than 15 years. These sites should be further investigated to verify if resurfacing was indeed applied after more than 15 years and to study the factors (e.g., preventive treatment) that led to their long longevity.
33

Measured Alligator Cracking (%)

30

25

20

15

10

I-520

5

0

0

5

State Route (LTPP)

10

Age (year)15

20

25

State Routes(non-LTPP) Interstate (LTPP) Interstate (non-LTPP)

Figure 4.2 Measured fatigue cracking

Rutting (in.)

0.6

0.5

0.4

0.3

0.2

0.1

0

0

5

10

15

Age (year)

State Route (LTPP) State Route (non-LTPP)l Interstate (LTPP)

20

25

Interstate (non-LTPP)

Figure 4.3 shows the measured rutting. Most of the measured rutting was between 0.05 in. and 0.35 in. Compared to the other non-interstate sites, the interstate sites exhibited moderate rutting (between 0.1in. and 0.3 in.).

34

Rutting (in.)

0.6

0.5

0.4

0.3

0.2

0.1

0

0

5

10

15

Age (year)

State Route (LTPP) State Route (non-LTPP)l Interstate (LTPP)

20

25

Interstate (non-LTPP)

Figure 4.3 Measured rutting Figure 4.4 shows the measured non-wheel path longitudinal cracking. There is dispersion in the

non-wheel path longitudinal cracking with a range from 0 to 12,000 ft/mile among the 38 sites.

The interstate sites, in general, exhibited minimum non-wheel path longitudinal cracking. Non-

wheel path longitudinal cracking was observed on the sites on I-520 in Richmond County and I-

95 in Chatham County. Again, some interstate sites show minimum longitudinal cracking after

more than 15 years of service. These sites should be further investigated to verify if resurfacing

was applied after more than 15 years and to study the factors that contribute to their long

longevity.

Measured Non-Wheelpath Longitudinal Cracking (ft/mile)

14000 12000 10000
8000 6000 4000 2000
0 0

I-520

5

10

15

Age (year)

20

25

All Interstate

35

Figure 4.4 Measured longitudinal cracking (non-wheel path) 4.1.2 Predicted Distresses This section compares the predicted and measured distresses on the interstate sites to verify the accuracy of the prediction models. Figure 4.5(a) shows the measured and predicted (at 50% reliability) fatigue cracking based on the data used in the calibration (ARA 2015a). The prediction is unbiased and reasonable with the data scattered around the equality line. Most of the measured and observed fatigue cracking was less than 2%. The only site with more fatigue cracking is on I-520. It is noted that for this site there was inconsistently in the predicted cracking reported in the study (ARA 2015a) and in the MEPDG file. While a 12.6% of the predicted fatigue was reported in the study, the MEPDG file output a lower cracking (3%). The MEPDG inputs and outputs should be further checked for future recalibration. Figure 4.5(b) shows the measured and predicted rutting on the interstate sites. Again, the data scattered around the equality line and the prediction was reasonable.

(a) Fatigue cracking

(b) Rutting

Figure 4.5 Measured vs. Predicted distresses (based on the data submitted by the ARA,
ARA 2015b) 36

4.2 Case Study on I-95 Site in Chatham County This section analyzes a typical interstate pavement structure designed in accordance with GDOT's current design procedures using the 1972 Design Guide and the MEPDG to provide some insight into the differences and their implications. The non-LTPP site on I-95 in Chatham County was selected as representative of GDOT's interstate pavement design (with four asphalt layers, OGFC, SMA, 19-mm binder layer, and 25-mm base layer on top of GAB). First, the pavement structures (e.g., as-built thickness) and actual traffic data collected on this section were analyzed using GDOT's pavement design tool, which was developed based on the 1972 Design Guide, to evaluate the design. Second, using the same data and the Georgia-calibrated distress transfer functions, the design was evaluated using the MEPDG. Finally, the optimized design to achieve the specified performance criteria suggested in GDOT's user guide was obtained using the new AASHTO Pavement ME Design software and compared to the current design.
The site on I-95 in Chatham County was originally constructed in 1964 with a 10 in., nondoweled jointed plain concrete pavement (JPCP) having a 30-ft joint spacing; it was later widened and overlaid using asphaltic concrete. In 1994, this section was widened with 8.5 in.of asphalt concrete layer on top of a 14 in. GAB. The top four layers were a 7/8 in. of open-graded layer (asphaltic concrete "D"), a 1.5 in. of dense-graded layer (asphaltic concrete "E"), a 2 in. of asphaltic concrete "B" (19 mm), and a 4 in. of 25-mm base layer, as shown in Figure 4.6. This section was resurfaced in 2002 to replace the open-graded layer. Since then, it has carried about 10 million heavy trucks over a 14-year time period. The one-way AADTT was 2,425 on the 6lane roadway (3 lanes in each direction), and the AADTT grew at a linear rate of 4.2%. The
37

subgrade was found to be an AASHTO A-2-4 soil with a resilient modulus of 16,500 psi based on the value used in the calibration.
Figure 4.6 Pavement Structure on I-95 Site in Chatham County 4.2.1 Analysis Using 1972 AASHTO Interim Design Guide In this section, the pavement structures and the traffic data on the selected site were analyzed using GDOT's pavement design tool, which is based on the 1972 Design Guide. The default soil support value (4) and regional factor (1.7) for Chatham County were used in the analysis. The percentages of single and multiple units were calculated based on the vehicle classification distribution of this section; the default single unit ESAL factor (0.4) and multiple unit ESAL factor (1.5) were used to calculate the design ESALs. Figure 4.7 shows the pavement design analysis generated by using GDOT's pavement design tool. To carry the 16.2 million heavy trucks for a 20-year design life, the required structure number is 5.70. The structure number based on the pavement structures on I-95 site is 5.12; thus, it was under-designed by 10.17%. It reached the required structure number at 8 million heavy trucks. To carry 16.2 million heavy trucks, the current pavement structure needs an additional 2 in. of asphalt concrete layer.
38

Figure 4.7 Pavement structure analysis using 1972 AASHTO Interim Design Guide
4.2.2 Analysis Using MEPDG The same pavement structure was analyzed using the MEPDG with Georgia's coefficients (ARA 2015a). As shown in Figure 4.8, this pavement structure can meet the performance criteria; all the predicted distresses at the specified reliability were lower than the threshold values at the end of the 20-year design life with accumulated 16 million heavy trucks. It is noted that a 95% reliability is used for fatigue cracking and rutting, as suggested in GDOT's user guide. When 50% reliability was used, the predicted distresses were much lower (0.25 in. of rutting, 0.62% of fatigue cracking). When the reliability increased from 50% to 95%, the predicted fatigue cracking significantly increased from 0.62% to 8.61%. This means the selection of reliability level has a big impact on the distress threshold values, which determines whether or not the pavement structure design passes the criteria. At 95% reliability, the MEPDG predicted an 8.61% of fatigue cracking, 0.35 in. of rutting, and 802 ft/mile of thermal cracking at the end of
39

20 years. The pavement structure can last longer than 20 years and can reach the performance criteria in 21 years with 16 million heavy trucks.
Figure 4.8 Pavement structure analysis using MEPDG Since the predicted distresses were below the performance criteria, an optimized redesign was performed to seek a pavement structure with thinner layers that met the performance criteria. Figure 4.9 shows the resulting design and summarizes the predicted distresses. The total HMA thickness for the 20-year design period is reduced by 0.5 inches. Thus, this new design costs less than the original design but still meets all the performance requirements.
40

Figure 4.9 Redesign of I-95 site Figure 4.10 shows the design thicknesses under different traffic using different methods. In general, the thicknesses designed by using the MEPDG are 0.5 in. to 1 in. less than the ones required by the 1972 Design Guide. However, it is noted that the OGFC was considered differently in these two methods. OGFC was not considered in the 1972 Design Guide because its structure coefficient was 0; it was considered in the MEPDG with reduced thickness (using a factor of 0.75). The more conservative design by using the 1972 Design Guide could allow GDOT to replace only the top porous friction course in 10-12 years and both the porous friction
41

course and SMA layer in 20-24 years without touching the underlying layers, which are protected by the extra pavement thickness and remain structurally sound.

Thickness of AC (in.)

14 12 10
8 6 4 2 0
1,919,900

4,686,600

7,161,300

Accumulated Trucks in Design Year

1972 Design MEPDG Design

11,066,800

Figure 4.10 AC thickness (1972 design vs. MEPDG)

4.3 Characterization of SMA Using GDOT's current 1972 Design Guide, both SMA and Superpave have a structure coefficient of 0.44, which means their contribution to the structure number is considered to be the same. As a result, the benefit of SMA (e.g., longer fatigue life, better rutting resistance and durability) cannot be counted during the design stage. In this section, we explore the characterization of SMA using the MEPDG and compare the predicted performance of SMA and Superpave. The principal mechanical property input for hot-mix asphalt in the MEPDG is dynamic modulus. The methods for specifying dynamic modulus at each of the three input levels in the MEDPG are as follows:
42

Level 1: Laboratory-measured dynamic modulus |E*| at multiple temperatures and loading frequencies (AASHTO TP62). In addition, binder stiffness and phase angle data are required for the global aging model.
Level 2: The Witczak |E*| predictive model is used to predict dynamic modulus based on gradation, volumetric data, binder stiffness, and phase angle data. The model used in the MEPDG is as follows:
Level 3: The Witczak |E*| predictive model is still used to predict E for Level 3. However, default binder stiffness and phase angle data are based on the default for the binder. The required data inputs are gradation and volumetric data.
The Witczak's prediction model is a purely empirical regression model developed from a large database of over 2700 laboratory test measurements. The databases used to develop and calibrate the Witczak's prediction model contain mostly conventional dense-graded mixtures but very few gap-graded SMA mixtures. Several studies (Ceylan et al., 2009; Cross et al., 2009; ODOT, 2009)
43

have found that the Witczak predictive model is dominated by temperature influences and does not do a good job of ranking mixtures in terms of their measured stiffness values at a given temperature and loading frequency. Especially, with higher AC content and a coarse stone skeleton, SMA has a lower E* compared to Superpave. Figure 4.11 shows the dynamic modulus estimated using the Witczak predictive model in the MEPDG based on the mix properties (e.g., gradation and AC content) suggested in GDOT's user guide. It shows the SMA has a slightly (~3%) lower dynamic modulus compared to Superpave. The difference in dynamic modulus is small, and it does not make much difference in the predicted distresses. However, this means that by using the Witczak predictive model (levels 2 and 3), the MEPDG cannot be used to justify the benefit of SMA in the design stage, either. Previous studies (e.g., Sotil et al., 2007; ODOT, 2009) have shown that, when tested without confinement, certain gap-graded mixtures (such as SMA mixtures) may have lower |E*| values than dense-graded mixtures. SMA mixtures may, therefore, show lower rutting resistance when modeled in the current MEPDG software, contrary to the observed superior rutting and cracking resistance of SMAs (Michael et al., 2003) in the field. A summary of states' studies on the dynamic modulus of SMA is in Appendix III.
Figure 4.11 E* for SMA and Superpave 44

45

5. CONCLUSIONS AND RECOMMENDATAIONS
The Georgia Department of Transportation (GDOT) is in the process of evaluating the use of the MEPDG for designing its new and rehabilitated pavement structures. GDOT has undertaken projects to establish the groundwork for the use of the MEPDG, including characterizing material properties, analyzing traffic loading, and calibrating the MEPDG performance prediction models to Georgia's conditions and materials. The GALTPP program was initiated by GDOT to provide sufficient sites for the initial MEPDG local calibration, and, more importantly, to continue long-term performance monitoring on the sites in which GDOT is interested to support performance evaluation and/or future MEPDG recalibration. The outcomes/findings will improve GDOT's practices on pavement design, material, construction, and maintenance. Currently, the GALTPP comprises 38 flexible pavement sites (17 LTPP and 21 non-LTPP sites) and 23 rigid pavement sites (11 LTPP and 12 non-LTPP sites). Various field and laboratory tests, including condition surveys in accordance with the LTPP Distress Identification Manual, FWD, and DCP for base and subgrade, bulk specific gravity measured on each layer, etc., were conducted on the non-LTPP sites, and documents (e.g., as-built plans and construction files) were gathered to provide the data needed for the calibration. These data are essential for further recalibration of the MEPDG, and additional data (e.g., performance data on current sites and data on new warm mix asphalt sites) are expected to be incorporated into the GALTPP program. Therefore, this project is designed to maintain the data collected for the GALTPP program, including the existing data and the data to be incorporated in the future. In addition, the project evaluates the MEPDG performance prediction for Georgia's interstate highways because they account for a major part of the
46

capital investments for roadways. The following are the major findings from Phase 1 of the project:
1) The data collected on GALTPP sites, including FWD, DCP, locations, etc., were gathered and carefully reviewed. The site locations were verified by comparing them (x-y coordinates included in GDOT's calibration study) to the core locations collected using GDOT's PDA-based core data collection application to ensure the sites can be correctly located for long-term performance monitoring. It was found that the site location data does not match the core location data. Thus, the site location data was corrected based on the first core located along the travel direction. The location data was further processed to obtain additional location information (e.g., RCLINK and milepoint) using GDOT's location reference system.
2) A database (GALTPP database) with location reference information was designed to store and manage the input parameters used in the MEPDG calibration, the condition survey data, the testing data, and the documents collected on the GALTPP sites. GALTPP database tables and fields for flexible pavement were designed based on a relational database concept with geospatial information so it can be integrated into a GIS (Geographic Information Systems) platform. The data was processed and populated into the GALTPP database. In addition, a GIS project, along with an add-in tool, was developed using the GALTPP database for visualizing the sites.
3) A review of GDOT's pavement condition survey data shows raveling is the predominate distress on Georgia's interstate pavements; 41% of interstate segments were reported with raveling in FY 2015. Raveling is also an important performance indicator that triggers the need for maintenance on the porous friction course (e.g., Open Graded Friction Course 47

(OGFC) or Porous European Mix (PEM)) on the surface layer, but it is not modeled in the MEPDG. 4) A total of 38 sites (17 LTPP and 21 non-LTPP sites) were used to calibrate the coefficients in the MEPDG transfer functions to eliminate bias and improve accuracy (i.e., reducing the standard error). Among them, five sites are on interstates. Compared to the other sites, these interstate sites exhibited low fatigue cracking (less than 3%) and moderate rutting (between 0.15 in and 0.3 in) at the end of pavement service interval (i.e., before pavement rehabilitation). The only site that exhibited more cracking (approximately 10% in 17 years) is on I-520, which has 7 in. of asphalt concrete layers. Based on the limited data, the measured distresses were within the distresses levels predicted at 50% reliability using the MEPDG. 5) A case study was conducted on I-95 in Chatham County based on the existing pavement structure. Using the MEPDG, the predicted distresses would reach the distress performance criteria (0.35" of rutting and 10% of fatigue cracking) at 95% reliability in 20 years. However, the observed distresses (0.25 in. of rutting and 3% of fatigue cracking) are close to the distresses predicted at 50% reliability. 6) Compared to the MEPDG, the current design procedure (1972 AASHTO Interim Design Guide is on the conservative side. The interstate pavement structure on the I-95 site is 10.17% under-designed when it was validated using the 1972 Design Guide. According to the 1972 Design Guide, to carry the 16.2 million heavy trucks, an additional 2 in. of asphalt base would be needed. However, the design without the 2 in. of asphalt base passed all the performance criteria when it was validated using the MEPDG.
48

7) Though it is on the conservative side, the current 1972 Design Guide allows GDOT to replace only the top porous friction course in 10 to 12 years, and both the porous friction course and SMA layer in 20 to 24 years without the need to replace the underlying layers because the underlying layers are still structurally sound with very limited distresses. Analyses on multiple pavement service intervals (e.g., more than 20 years) based on GDOT's maintenance practices would help to determine the most cost-effective pavement structures.
8) Based on the field observation, SMA performs better in term of fatigue life on heavily traveled roads than does Superpave. This benefit is not modeled in the current design procedure (the 1972 Design Guide) because both SMA and Superpave have the same structure coefficient of 0.44. This issue remains the same in the MEPDG. Moreover, using the Witczak predictive model in the MEPDG, the SMA has a slightly lower dynamic modulus than Superpave. This leads to higher predicted rutting and fatigue cracking, although the differences are small. However, this is contrary to the observed field performance in Georgia, which shows better rutting and fatigue resistance of SMA.
To move forward in maintaining an active GALTPP program, the following are recommended: 1) It is recommended that distress and FWD data be collected on an annual or biennial basis on
the GALTPP sites as the pavements continue deteriorating to establish a long-term performance monitoring. Especially, it is recommended cracking data be collected before and after resurfacing on I-95 site in Chatham County, which will be resurfaced next year. Such data will allow GDOT to validate the performance on this site and assess the development of cracking on micro-milled surfaces. 2) The 3D laser technology (e.g., 3D pavement data, video log images, etc.) can be used for collecting consistent and detailed pavement distress data on the GALTPP sites. The high-
49

resolution 3D laser data can be used for detecting cracks and quantitatively and objectively measuring raveling on the porous friction course surfaces. In addition, it can collect the cracking data before and after micro-milling is performed on the porous friction course. These data are invaluable for assessing the development of top-down and bottom-up cracking on Georgia's pavements. 3) A raveling prediction model (including a measure for quantifying raveling) can be developed and incorporated into the life-cycle analysis of the interstate pavement design (for new and rehabilitated pavement structures). This allows GDOT to reliably quantify raveling and identify the timing for adequate treatment(s), which is difficult using current visual inspection methods. 4) It is recommended that life-cycle cost analysis (based on GDOT's maintenance practices) be conducted to determine the pavement structure design that is most cost-effective for a pavement's full life cycle.
50

51

REFERENCES
(1) AASHTO. 1972. AASHTO Guide for the Design of Pavement Structures.ARA. 2015a. Calibration of the MEPDG Transfer Functions in Georgia. Georgia Department of Transportation, Atlanta, GA.
(2) ARA. 2015b. Implementation of the Mechanistic-Empirical Pavement Design Guide in Georgia. FHWA - GA -14 - 11 17, Georgia Department of Transportation, Atlanta, GA.
(3) AASHTO. 2010. Guide for the Local Calibration of the Mechanistic-Empirical Pavement Design Guide. American Association of State Highway and Transportation Officials (AASHTO), Washington, DC.
(4) Barksdale, Richard. GDOT Research Project 9217: Optimum Design of Stone Matrix Asphalt Mixes. Georgia Department of Transportation, 1995.
(5) Craus, J., A. Chen, J. Sousa, and C. Monismith. Development of Failure Curves and Investigation of Asphalt Concrete Pavement Cracking from Super-Overloaded Vehicles. Division of New Technology, Materials, and Research, California Department of Transportation, Sacramento, 1994.
(6) Dauzats, M., and A. Rampal. Mechanism of Surface Cracking in Wearing Courses. Proc., 6th International Conference on Structural Design of Asphalt Pavements, University of Michigan, Ann Arbor, 1987, pp. 232247.
(7) GDOT. 1993. Georgia's Concrete Pavement Condition Survey Instruction Manual. Georgia Department of Transportation, Atlanta, GA.
(8) Jared, David. GDOT Research Project 9102: Evaluation of Stone Matrix Asphalt and Porous European Mix. Georgia Department of Transportation, 1997.
(9) Jared, David. GDOT Research Project 9202: Evaluation of Stone Matrix Asphalt Overlay on Portland Cement Concrete. Georgia Department of Transportation, 1997.
(10) Lai, J., M. Bruce, D. M. Jared, P. Y. Wu, and S. Hines. (2012). Pavement Preservation with Micro-milling in Georgia Follow Up Study. Journal of The Transportation Research Record, No.2292, pp.81-87.
52

(11) Matsuno, S., and T. Nishizawa. Mechanisms of Surface-Initiated Longitudinal Wheel Path Cracks in High-Type Bituminous Pavements. Proc., 7th International Conference on Asphalt Pavements, Vol. 2, University of Nottingham, 1992, pp. 277291.
(12) NCHRP. 2004. Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures (NCHRP Project 1-37A). Online Version (http://www.trb.org/mepdg).
(13) Tsai, Y. and Wu, Y. Developing A Georgia Long-Term Pavement Performance (GALTPP) Monitoring Program Design And Site Selection. Georgia Department of Transportation, Atlanta, GA, 2014.
(14) Tsai, Y., Wu, Y., Lai, J., Geary, G. (2012). Ridge-to-Valley Depth Measured with Road Profiler to Control Micromilled Pavement Textures for Super-Thin Resurfacing on I-95, Journal of The Transportation Research Record, No.2306, pp.144-150.
(15) Tsai, Y., Wu, Y., Lewis, Z. (2014) "Full-Lane Coverage Micromilling Pavement-Surface Quality Control Using Emerging 3D Line Laser Imaging Technology" Journal of Transportation Engineering, Vol. 140(2)
(16) Tsai, Y. and Wu, Y. (2015) "Sustainable and Cost-Effective Pavement Preservation Method: Micro-Milling and Thin Overlay" Journal of Transportation Engineering, (in press)
(17) Wang, Z. and Tsai, Y. Enhancement of GDOT's Pavement Rehabilitation and Design Processes by Integrating New and Existing Data Sources and Developing Data Analysis. 2013. GDOT, Atlanta, GA.
(18) Watson, Donald, and Jared, David. Stone Matrix Asphalt: Georgia's Experience. Georgia Department of Transportation, 1995.
(19) Uhlmeyer, J. S., K. Willoughby, L.M. Pierce and J.P. Mahoney (2000), "Top-Down Cracking in Washington State Asphalt Concrete Wearing Courses," Journal of the Transportation Research Board, No. 1730, TRB, Washington, D.C., pp. 110-116.
53

APPENDIX A: SITE LOCATION

Site

Type

Widening, Mill, F1 & Overlay

F2 Mill & Overlay

Widening & F3 Overlay

F4 Mill & Overlay

F7 Overlay F8 Mill & Inlay

County Banks & Jackson Cobb
Jackson Douglas Fulton & Clayton Clayton

F9 Mill & Overlay Fulton

Widening &

F10 Overlay

Bryan

Semi-Rigid;

F11 Overlay

Decatur

Widening

Chatham &

F12 &Overlay

Effingham

Reconstruction, Thomas &

F13 Semi-Rigid

Brooks

F16 Overlay

Mitchell

Wilkinson &

F18 Mill & Overlay Washington

Reconstruction,

F19 Conventional Richmond

Reconstruction,

F20 Conventional Polk

Reconstruction,

F21 Conventional Cherokee

Reconstruction,

F23 Full-Depth

Jefferson

F24 Overlay

Oconee

Widening &

F25 Overlay

Pike

Reconstruction,

F26 Conventional Worth

Route

X

SR-15
SR-180 SR-11/ US-129
SR-6

-83.45407 -84.50888
-83.68865 -84.63973

SR-6
SR-54 I-85/ SR-403

-84.53662 -84.34891
-84.48281

SR-144 SR-1/ US-27

-81.32511 -84.55647

I-95/I-16 SR38/US-84 SR-3/ US-19

-81.23941 -83.77603 -84.08231

SR-57

-83.12303

I-520 SR-6/ US-278

-81.97540 -85.26138

SR-108 -84.58756

SR-171 SR-24/ US-129

-82.46211 -83.42801

SR-109 -84.31036

SR-256 -83.79793

Y

RCLINK

34.23695 1571001500 33.85098 0671028000

34.17628 1571033200 33.79584 0971000600

33.66042 1211000600 33.54670 0631005400

33.61924 1211040300

31.95789 0291014400

30.86661 0871000100

32.08687 0511040500

30.80000 2751003800

31.08023 2051000300

32.81008 3191005700

33.41180 2451041500

34.00514 2331000600

34.26446 0571010800

33.01921 1631017100

33.77210 2191002400

33.04869 2311010900

31.49693 3211025600

MP 16.30 2.94 10.40 1.03 6.70 5.95 11.84 9.15 16.84 8.39 23.68 0.22 17.29 10.81 9.28 3.98 18.19 3.92 15.07 14.26

A-1

APPENDIX B: GALTPP TABLES

Section information

Field Name GALTPP_SEC_CON_ID GALTPP_SECTION_ID CONSTRUCTION_ID LTPP_SECTION_ID COUNTY ROUTENO ROUTE_SUFFIX Milepoint_FROM Milepoint_TO Milepost_FROM Milepost_TO

Units

DIRECTION_OF_TRAVEL

LANE_NUMBER

FUNCTIONAL_CLASS

TOT_LANES

PAVEMENT_TYPE

LANE_WIDTH

ft

SHOULDER_TYPE

SHOULDER_WIDTH

ft

DIVIDED

DATE_EARTHWORK

DATE_HMA_PLACED TRAFFIC_OPEN_DATE

Field Type CHARACTER CHARACTER CHARACTER CHARACTER(6)

Description An identification number GALTPP_SECTION_ID+ CONSTRUCTION _ID. Test section identification number (one for each section).
Construction event in sequence.
LTPP test section identification.

CHARACTER(3) CHARACTER(4) CHARACTER(2) NUMBER NUMBER

County in which the test section is located.
The route number for the route that the section is located on. The route suffix for the route that the section is located on. Beginning mile point
Ending mile point

NUMBER

Beginning mile post for interstate highways

NUMBER CHARACTER(1)
NUMBER(1,0) CHARACTER

Ending mile post for interstate highways
E for East, W for West, N for North, S for South base on the direction of travel within the lane for which data is being collected.
The number of the lane on which data is being collected. 1 is the outside lane. The others are numbered consecutively as you move to the inside edge of the pavement.
Functional class of roadway on which section is located.

NUMBER(1,0)

Total number of lanes in one direction.

CHARACTER

NUMBER(2,0) CHARACTER(7) NUMBER(2,0) CHARACTER(1) DATE

Width of the lane the test section occupies.
Indication of whether the shoulder is "paved," "unpaved," or "none."
The width of the shoulder in feet.
Y or N indicating that the roadway does or does not have a median.
Date the earthwork was completed in the construction of the project.

DATE

Date the hot-mix asphalt was placed in the construction of the project.

DATE

Date the test section was opened to traffic.
B-1

Field Name LATITUDE LONGITUDE ELEVATION
LOCATION_INFO

Units Field Type Degrees NUMBER(5,3)

Description Latitude of the test section in degrees.

Degrees NUMBER(5,3)

Ft

NUMBER(4,0)

Longitude of the test section in degrees.
Estimate of the elevation of the test section relative to sea level.

CHARACTER(100) Description of the location of the test section.

Performance information

This table stores the distress inputs for the MEPDG models for flexible pavement.

Field Name

Units

Field Type

Description

GALTPP_SEC_CON_ID

CHARACTER

A unique identifier for GALTPP

SOURCE SURVEY_DATE

SURFACE_DOWN_FATIGUE %

BOTTOM_UP_CRACKING

%

THERMAL_CRACK

ft/mi

AVG_WIRELINE_RUT_DEPTH in

STD_WIRELINE_RUT_DEPTH in

STUDDED_TIRE_WEAR

in

CHARACTER DATE NUMBER(3,1) NUMBER(3,1)) NUMBER(4,1) NUMBER(3,2) NUMBER(3,2) NUMBER(3,2)

Source of the distress data (COPACES, LTPP)
Date of distress survey.
Percentage of wheel path area that has experienced surface-down fatigue. Percentage of wheel path area that has experienced bottom-up cracking. Total length of thermal cracking per lane-mile.
Average rut depth for the 500-ft test section.
Standard deviation of rut depth measurements taken on the test section. Portion of rut depth due to wearing of the surface from studded tires.

This table stores the distress data for flexible pavement.

Field Name

Units Field Type

Description

GALTPP_SEC_CON_ID

CHARACTER Test section identification number.

SOURCE SURVEY_DATE GATOR_CRACK_A_L GATOR_CRACK_A_M GATOR_CRACK_A_H BLK_CRACK_A_L BLK_CRACK_A_M BLK_CRACK_A_H
EDGE_CRACK_L_L
EDGE_CRACK_L_M

CHARACTER Source of the distress data (COPACES, LTPP)

DATE

(mm/dd/yyyyhh Date of distress survey.

:mi:s)

ft2

NUMBER(5,1) Area of alligator (fatigue) cracking of low severity

ft2

NUMBER(5,1)

Area of alligator (fatigue) cracking of moderate severity may be evident).

ft2

NUMBER(5,1)

Area of alligator (fatigue) cracking of high severity may be evident).

ft2

NUMBER(5,1) Area of block cracking of low severity

ft2

NUMBER(5,1) Area of block cracking of moderate severity

Area of high severity block cracking (mean crack width

ft2

NUMBER(5,1) greater than 19 mm or under 19 mm with moderate to

high severity random cracking).

ft

NUMBER(4,1)

Length of low severity edge cracking (cracks without break up or loss of material).

Length of moderate severity edge cracking (cracks with

ft

NUMBER(4,1) some break up and loss of material for up to 10 percent

of the affected length).

B-2

Field Name EDGE_CRACK_L_H LONG_CRACK_WP_L_L
LONG_CRACK_WP_L_M
LONG_CRACK_WP_L_H
LONG_CRACK_WP_SEAL_L_L
LONG_CRACK_WP_SEAL_L_M
LONG_CRACK_WP_SEAL_L_H
LONG_CRACK_NWP_L_L
LONG_CRACK_NWP_L_M
LONG_CRACK_NWP_L_H LONG_CRACK_NWP_SEAL_L_ L LONG_CRACK_NWP_SEAL_L_ M LONG_CRACK_NWP_SEAL_L_ H REFL_CRACK_TRANS_NO_L REFL_CRACK_TRANS_NO_M
REFL_CRACK_TRANS_NO_H REFL_CRACK_TRANS_L_L

Units ft ft ft ft ft ft ft ft ft ft ft ft ft
ft

Field Type NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(3,0) NUMBER(3,0) NUMBER(3,0) NUMBER(5,1)
B-3

Description
Length of high severity edge cracking (considerable break up and loss of material for more than 10 percent of the affected length). Length of low severity, longitudinal cracking in wheel path (cracks of unknown width well sealed or with mean width of 6 mm or less). Length of moderate severity, longitudinal cracking in wheel path (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of high severity, longitudinal cracking in wheel path (mean crack width greater than 19 mm or under 19 mm with adjacent moderate to high severity random cracking). Length of low severity, well-sealed longitudinal cracking in wheel path (cracks of unknown width or with mean width of 6 mm or less). Length of moderate severity, well-sealed longitudinal cracking in wheel path (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of high severity, well-sealed longitudinal cracking in wheel path (crack mean width greater than 19 mm or under 19 mm with adjacent moderate to high severity random cracking). Length of low severity, non-wheel path longitudinal cracking (cracks of unknown width well sealed or with mean width of 6 mm or less). Length of moderate severity, non-wheel path longitudinal cracking (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of high severity, non-wheel path longitudinal cracking (mean crack width greater than 19 mm or fewer than 19 mm with adjacent moderate to high severity random cracking). Length of low severity, well-sealed non-wheel path longitudinal cracking (cracks of unknown width or with mean width of 6 mm or less). Length of moderate severity, well-sealed non- wheel path longitudinal cracking (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of high severity, well-sealed non-wheel path longitudinal cracking (mean crack width greater than 19 mm or fewer than 19 mm with adjacent moderate to high severity random cracking). Number of low severity, transverse reflection cracks (cracks of unknown width well sealed or with mean width of 6 mm or less). Number of moderate severity, transverse reflection cracks (mean crack width of 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Number of high severity, transverse reflection cracks (mean crack width greater than 19 mm or under 19 mm with adjacent moderate to high severity random cracking). Length of low severity, transverse reflection cracking at joints (cracks of unknown width well sealed or with

Field Name
REFL_CRACK_TRANS_L_M
REFL_CRACK_TRANS_L_H
REFL_CRACK_TRANS_SEAL_ L_L REFL_CRACK_TRANS_SEAL_ L_M REFL_CRACK_TRANS_SEAL_ L_H
REFL_CRACK_LONG_L_L
REFL_CRACK_LONG_L_M
REFL_CRACK_LONG_L_H
REFL_CRACK_LONG_SEAL_L _L REFL_CRACK_LONG_SEAL_L _M
REFL_CRACK_LONG_SEAL_L _H
TRANS_CRACK_NO_L
TRANS_CRACK_NO_M
TRANS_CRACK_NO_H
TRANS_CRACK_L_L
TRANS_CRACK_L_M TRANS_CRACK_L_H

Units ft ft ft ft ft ft ft ft ft ft ft
ft ft ft

Field Type
NUMBER(5,1) NUMBER(5,1) NUMBER(5,1) NUMBER(5,1) NUMBER(5,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(4,1) NUMBER(3,0) NUMBER(3,0) NUMBER(3,0) NUMBER(5,1) NUMBER(5,1) NUMBER(5,1)

Description
mean width of 6 mm or less).
Length of moderate severity, transverse reflection cracking at joints (mean crack width of 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of high severity, transverse reflection cracking at joints (mean crack width greater than 19 mm or fewer than 19 mm with adjacent moderate to high severity random cracking). Length of well-sealed, low severity transverse cracking (cracks of unknown width or with mean width of 6 mm or less). Length of well-sealed, moderate severity transverse cracking (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of well-sealed, high severity transverse cracking (mean crack width greater than 19 mm or under 19 mm with adjacent moderate to high severity random cracking). Length of low severity, longitudinal reflection cracking at joints (cracks of unknown width well sealed or with mean width of 6 mm or less). Length of moderate severity, longitudinal reflection cracking at joints (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of high severity, longitudinal reflection cracking at joints (mean crack width greater than 19 mm or fewer than 19 mm with adjacent moderate to high severity random cracking). The length of well-sealed, low severity longitudinal reflection cracking at joints (cracks of unknown width or with mean width of 6 mm or less). The length of well-sealed, moderate severity longitudinal reflection cracking at joints (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). The length of well-sealed, high severity longitudinal reflection cracking at joints (mean crack width greater than 19 mm or under 19 mm with adjacent moderate to high severity random cracking). Number of low severity transverse cracks (cracks of unknown width well sealed or with mean width of 6 mm or less). Number of moderate severity transverse cracks (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Number of high severity transverse cracks (mean crack width greater than 19 mm or under 19 mm with adjacent moderate to high severity random cracking). Length of low severity transverse cracking (cracks of unknown width well sealed or with mean width of 6 mm or less). Length of moderate severity transverse cracking (crack mean width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). Length of high severity transverse cracking (mean crack width greater than 19 mm or under 19 mm with adjacent

B-4

Field Name
TRANS_CRACK_SEAL_L_L
TRANS_CRACK_SEAL_L_M
TRANS_CRACK_SEAL_L_H
PATCH_NO_L PATCH_NO_M PATCH_NO_H PATCH_A_L PATCH_A_M PATCH_A_H POTHOLES_NO_L POTHOLES_NO_M POTHOLES_NO_H POTHOLES_A_L POTHOLES_A_M POTHOLES_A_H SHOVING_NO SHOVING_A BLEEDING POLISH_AGG_A RAVELING PUMPING_NO PUMPING_L OTHER

Units ft ft ft
ft2 ft2 ft2
ft2 ft2 ft2 ft2 ft2 ft2 ft2
ft

Field Type
NUMBER(5,1)
NUMBER(5,1)
NUMBER(5,1)
NUMBER(3,0) NUMBER(3,0) NUMBER(3,0) NUMBER(5,1) NUMBER(5,1) NUMBER(5,1) NUMBER(3,0) NUMBER(3,0) NUMBER(3,0) NUMBER(5,1) NUMBER(5,1) NUMBER(5,1) NUMBER(3,0) NUMBER(5,1) NUMBER(5,1) NUMBER(5,1)
NUMBER(5,1) NUMBER(3,0) NUMBER(4,1) CHARACTER( 80)

Description
moderate to high severity random cracking).
The length of well-sealed, low severity transverse cracking (cracks of unknown width or with mean width of 6 mm or less). The length of well-sealed, moderate severity transverse cracking (mean crack width from 6 to 19 mm or under 19 mm with adjacent low severity random cracking). The length of well-sealed, high severity transverse cracking (mean crack width greater than 19 mm or under 19 mm with adjacent moderate to high severity random cracking). Number of patches/patch deteriorations with low severity distress of any type. Number of patches/patch deteriorations with moderate severity distress type. Number of patches/patch deteriorations with high severity distress of any type. Area of patching with low severity distress or patch deterioration. Area of patching with moderate severity distress or patch deterioration. Area of patching with high severity distress or patch deterioration. Number of low severity potholes (less than 25 mm deep). Number of moderate severity potholes (from 25 to 50 mm deep). Number of high severity potholes (more than 50 mm deep).
Area of low severity potholes (less than 25 mm deep).
Area of moderate severity potholes (from 25 to 50 mm deep).
Area of high severity potholes (more than 50 mm deep).
Number of areas where shoving exists.
The area of shoving, localized longitudinal displacement of the pavement surface. Presence of excess asphalt on the pavement surface, which may create a shiny, glass-like reflective surface. Area of polished aggregate (binder worn away to expose coarse aggregate). Wearing away of the pavement surface caused by the dislodging of aggregate particles and loss of asphalt binder.
Number of occurrences of water bleeding and pumping.
Length of pavement affected by water bleeding and pumping.
A description of other surface distress.

This table stores the distress inputs for the MEPDG models for JPCP.

Field Name

Units

Field Type

Description

GALTPP_SEC_CON_ID

CHARACTER

A unique identifier for GALTPP

LTPP_SECTION_ID

CHARACTER(6) LTPP test section identification.

B-5

Field Name SURVEY_DATE FAULTING CRACKING

Units
in %

Field Type DATE NUMBER(3,1) NUMBER(3,1)

Description Date of distress survey. Mean joint faulting % slabs cracked

This table stores the distress data for JPCP.

Field Name GALTPP_SEC_CON_ID SOURCE SURVEY_DATE SURVEYOR BEFORE_TEMP AFTER_TEMP AVG_FAULTING MIN_FAULTING MAX_FAULTING STD_FAULTING BROKEN_SLABS
CORNER_BREAKS_NO_L
CORNER_BREAKS_NO_M
CORNER_BREAKS_NO_H LONG_CRACK_L_L
LONG_CRACK_L_M
LONG_CRACK_L_H LONG_CRACK_SEAL_L_L LONG_CRACK_SEAL_L_M
LONG_CRACK_SEAL_L_H TRANS_CRACK_NO_L TRANS_CRACK_NO_M TRANS_CRACK_NO_H TRANS_CRACK_L_L

Field Type CHARACTER CHARACTER
DATE CHARACTER
NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER
NUMBER
NUMBER
NUMBER NUMBER
NUMBER
NUMBER NUMBER NUMBER
NUMBER NUMBER NUMBER NUMBER NUMBER

Description A unique identifier for GALTPP
Source of the distress data (COPACES, LTPP).
Date survey was performed.
Person who conducts the survey.
Pavement surface temperature at the beginning of the distress survey.
Pavement surface temperature at the end of the distress survey.
Average edge faulting calculated per site per survey.
Minimum edge faulting per site per survey.
Maximum edge faulting per site per survey.
Standard deviation for edge faulting calculated per site per survey.
Total number of broken slabs.
Number of low severity corner breaks. (Notspalled for more than 10 percent of length; no measurable faulting; corner piece not broken in two or more pieces.) Number of moderate severity corner breaks. (Spalled at low severity for more than 10 percent; or faulting less than 13 mm; corner piece not broken in two or more Number of high severity corner breaks. (Spalled at moderate to high severity for more than 10 percent of crack; or faulting exceeds 13 mm or corner piece in two or more pieces.) Length of low severity longitudinal cracking. (Crack widths less than 3 mm, no spalling or measurable faulting.) Length of well-sealed, moderate severity longitudinal cracking. (Crack widths between 3 and 13 mm or spalling less than 75 mm or faulting up to 13 mm.) Length of high severity longitudinal cracking. (Crack widths greater than 13 mm or spalling greater than 75 mm or faulting greater than 13 mm.) Length of well-sealed, low severity longitudinal cracking. (Crack widths less than 3 mm, no spalling or measurable faulting.) Number of transverse cracks for which moderate severity distress is the highest level observed for at least 10 percent of the crack. Length of well-sealed, high severity longitudinal cracking. (Crack widths greater than 13 mm or spalling greater than 75 mm or faulting greater than 13 mm.) Number of low severity transverse cracks. (No spalling exceeding 10 percent of length). Number of transverse cracks for which moderate severity distress is the highest level observed for at least 10 Number of transverse cracks for which high severity distress exceeds 10 percent of the length. Length of low severity transverse cracking. (Crack widths less than 3 mm, no spalling and no measurable faulting.)
B-6

Field Name
TRANS_CRACK_L_M TRANS_CRACK_L_H LONG_SPALLING_L_L
LONG_SPALLING_L_M LONG_SPALLING_L_H TRANS_SPALLING_NO_L TRANS_SPALLING_NO_M TRANS_SPALLING_NO_H TRANS_SPALLING_L_L TRANS_SPALLING_L_M TRANS_SPALLING_L_H SCALING_NO SCALING_A POLISH_AGG_A BLOWUPS_NO PATCH_FLEX_NO_L PATCH_FLEX_NO_M PATCH_FLEX_NO_H PATCH_FLEX_A_L PATCH_FLEX_A_M PATCH_FLEX_A_H PATCH_RIGID_NO_L PATCH_RIGID_NO_M PATCH_RIGID_NO_H PATCH_RIGID_A_L PATCH_RIGID_A_M PATCH_RIGID_A_H PUMPING_NO PUMPING_L

Field Type
NUMBER NUMBER NUMBER
NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER NUMBER

Description
Length of moderate severity transverse cracking. (Crack widths between 3 and 6 mm or spalling fewer than 75 mm or faulting up to 6 mm.) Length of high severity transverse cracking. (Crack widths greater than 6 mm or spalling over 75 mm or faulting over 6 mm.) Length of low severity spalling of longitudinal joints. (Spalls less than 75 mm measured to center of joint with no loss of material.) Length of moderate severity spalling of longitudinal joints. (Spalls between 75 and 150 mm wide measured to center of joint with loss of material.) Length of high severity spalling of longitudinal joints. (Spalls greater than 150 mm measured to center of joint with loss of material.) Number of transverse joints with low severity spalling. (Spalls less than 75 mm wide measured to center of joint.) Number of transverse joints with moderate severity spalling. (Spalls between 75 and 150 mm wide measured to center of joint.) Number of transverse joints with high severity spalling. (Spalls more than 150 mm wide measured to center of joint.) Length of low severity spalling of transverse joints. (Spalls less than 75 mm measured to center of joint or with no loss of material.) Length of moderate severity spalling of transverse joints. (Spalls 75 to 150 mm wide measured to center of joint with loss of material). Length of high severity spalling of transverse joints. (Spalls more than 150 mm wide measured to center of joint with loss of material.) Number of areas with scaling.
Area of scaling (Deterioration of upper slab surface between 3 and 13 mm). Area of polished aggregate (Surface worn away to expose coarse aggregate). Number of blowups.
Number of flexible patches showing at most low severity distress of any type and no settlement at the perimeter. Number of flexible patches showing moderate severity distress of any type or settlement of up to 6 mm at the perimeter. Number of flexible patches showing high severity distress or settlement of 6 mm or more at the perimeter. Area of flexible patching showing, at most, low severity distress of any type and no settlement at the perimeter. Area of flexible patching showing moderate severity distress of any type or settlement of up to 6 mm at the perimeter. Area of flexible patching showing high severity distress of any type or settlement of 6 mm or more at the perimeter. Number of rigid patches showing, at most, low severity distress of any type and no settlement at the perimeter. Number of rigid patches showing moderate severity distress of any type or settlement of up to 6 mm at the perimeter. Number of rigid patches showing high severity distress of any type or settlement of 6 mm or more at the perimeter. Area of rigid patching showing, at most, low severity distress of any type and no settlement at the perimeter. Area of rigid patching showing moderate severity distress of any type or settlement of up to 6 mm at the perimeter. Area of rigid patching showing high severity distress of any type or settlement of 6 mm or more at the perimeter. Number of occurrences of water bleeding and pumping.
Length of pavement affected by water bleeding and pumping.
B-7

Traffic Inputs (HIF-11-026) This table includes a description of each traffic data element.

Name GALTPP_SEC_CON_ID

Description A unique identifier for GALTPP

AADTT

Initial two-way average annual daily truck traffic

Direction

Direction of traffic

No_Design_Lane

Number of lanes in the design direction

%_Trcks_Dsgn_Dir

Percent of trucks in the design direction (%)

%_Trcks_Dsgn_Lane

Percent of trucks in design lane (%)

Speed

Operational speed (mph)

Growth_Rate

Traffic growth rate (%)

General Traffic Inputs

Wheel_Location

Mean wheel location (inches from the lane marking)

Trffc_Wander_Stdev

Traffic wander standard deviation (in)

Design_Lane_Width

Design lane width (ft)

Axle Configuration

Avg_Axle_Width

Average axle width (edge-to-edge), outside dimension (ft)

Dual_Tire_Spacing

Dual tire spacing (in)

Tire_Pressure

Tire pressure (psi)

Axle_Spcing_Tandem

Tandem axle spacing (in)

Axle_Spcing_Tridem

Tridem axle spacing (in)

Axle_Spcing_Quad

Quad axle spacing (in)

Wheelbase

Wheelbase_Short

Average short axle spacing (ft)

% Trucks_Short

Percent of trucks short axle spacing (%)

Wheelbase_Medium

Average medium axle spacing (ft)

% Trucks_Medium

Percent of trucks medium axle spacing (%)

Wheelbase_Long

Average long axle spacing (ft)

% Trucks_Long

Percent of trucks long axle spacing (%)

Axle/Truck

Number of axles/truck

Class

FHWA truck class 4 13

Single

Average number of single axles per truck class

Tandem

Average number of tandem axles per truck class

Tridem

Average number of tridem axles per truck class

Quad

Average number of quad axles per truck class

Traffic Volume Adjustment Factors

Hour Distrib

Hourly distribution

Midnight 11:00 PM

Hourly truck traffic distribution by hour (%)

B-8

Total

Name

Description Sum of hourly distribution (must total 100%)

Monthly Adjust

Monthly adjustments

Month

Month of the year (January December)

Class_1 Class_13

Monthly adjustment factor for each FHWA truck class 1 13

Vehicle Distrib

Vehicle class distribution

Class_1 Class_13

AADTT distribution by vehicle class (%)

Total

Sum of AADTT distribution (must total 100%)

Axle Load Distribution Factors

Single

Single axle

Month

Month of the year (January December)

Class

FHWA truck class 1 13

Total

Sum of axle load distribution factors (must total 100%)

3000 41000

Percent of axles in each load interval (1000 lb increments)

Tandem

Tandem axle

Month

Month of the year (January December)

Class

FHWA truck class 1 13

Total

Sum of axle load distribution factors (must total 100%)

6000 82000

Percent of axles in each load interval (2000 lb increments)

Tridem

Tridem axle

Month

Month of the year (January December)

Class

FHWA truck class 1 13

Total

Sum of axle load distribution factors (must total 100%)

12000 102000

Percent of axles in each load interval (3000 lb increments)

Quad

Quad axle

Month

Month of the year (January December)

Class

FHWA truck class 1 13

Total

Sum of axle load distribution factors (must total 100%)

12000 102000

Percent of axles in each load interval (3000 lb increments)

Materials Inputs (HIF-11-026): This table includes a description of each AC material data element.

Field Name

Description

GALTPP_SEC_CON_ID LAYER_NO DESCRIPTION

Test section identification number.
Unique sequential number assigned to pavement layers, starting with layer 1 as the deepest layer (subgrade).
Code indicating general type of layer.

LAYER_TYPE

A character code indicating the type of layer.

LAYER_THICKNESS MATERIAL

Thickness of the layer. Code indicating the material used in the layer.

B-9

This table includes a description of each AC material data element.

Name

Description

GALTPP_SEC_CON_ID Test section identification number.

LAYER_NO

Layer number

Effctv_Bndr_Cntnt

Effective binder content (by weight)

Poisson_Ratio

Poisson's ratio

Existing_Layer

Existing layer as opposed to a new layer

Layer_Thickness

Layer thickness (in)

Air_Voids

Percent air voids

Thermal_Cndctvy

Thermal conductivity. (BTU/hr-ft-F)

Ref_Temp

Reference temperature (F)

Unit_Weight

Total unit weight (pcf)

Heat_Capacity

Heat capacity (BTU/lb-F)

E*

Dynamic modulus of asphalt mixture (Level 1)

Temperature

Temperature (F)

E*_0_1

Dynamic modulus (psi) at 0.1 Hz

E*_1

Dynamic modulus (psi) at 1 Hz

E*_10

Dynamic modulus (psi) at 10 Hz

E*_25

Dynamic modulus (psi) at 25 Hz

RTFO_SP

Superpave binder test data (Level 1 and Level 2)

Temperature

Temperature (F)

G*

Binder dynamic modulus (Pa)

Delta

Phase angle

RTFO_Conv

Conventional binder properties (Level 1 and Level 2)

Temp

Temperature (F)

Softening_Pnt

Softening point (P)

Abslt_Vscsty

Absolute viscosity (P)

Knmtc_Vscsty

Kinematic viscosity (CS)

Spcfc_Grvty

Specific gravity

Penetration

Penetration

Brkfld_Vscsty

Brookfield viscosity

Gradation

Gradation properties of asphalt mixture (Level 2 and Level 3)

Retained_3/4

Cumulative percent retained on the in sieve.

Retained_3/8

Cumulative percent retained on the in sieve.

B-10

Name Retained_ No_4 Passing_No_200 Creep Load_Time Creep_-4F Creep_-14F Creep_-32F Binder Binder_Type Binder_Grad ThermCrk Tnsl_Strngth VMA Aggrgt_CTC Mix_CTC

Description Cumulative percent retained on the #4 sieve. Percent passing the No. 200 sieve. Creep compliance properties (thermal cracking). Loading time (sec). Low temperature (-4 F). Mid temperature (14 F). High temperature (32 F). Asphalt binder properties (Level 3). Binder Type Binder grade Thermal cracking properties Average tensile strength at 14 F (psi) Mixture voids in mineral aggregate (%) Aggregate coefficient of thermal contraction (in/in/F) Mix coefficient of thermal contraction (in/in/F)

This table includes a description of each PCC material data element.

Name

Description

GALTPP_SEC_CON_ID LAYER_NO

Test section identification number. Layer number

CTE

Coefficient of thermal expansion (per F x 10-6)

Existing_Layer

Existing layer as opposed to a new layer

Unit_Weight

Unit weight (pcf)

Therm_Conduct

Thermal conductivity (BTU/hr-ft-F)

Poisson_Ratio

Poisson's ratio

Heat_Capacity

Heat capacity (BTU/lb-F)

Design

Concrete pavement design features

Curl/Warp_Effective_ Temperature_Difference Joint_Spacing

Permanent curl/warp effective temperature difference (F) Joint spacing (ft)

Sealant_Type

Joint sealant type

Dowel_Diameter

Dowel bar diameter (in)

Dowel_Spacing

Dowel bar spacing (in)

Tied_PCC

Identifies the presence of a tied concrete shoulder

B-11

Tied_LTE

Name

Widened_Slab

Slab_Width

PCC-Base_Interface

Base_Erodobility_Index

Loss_of_Friction

Steel_Reinforcement

Reinforcement_Steel_Diameter

Depth_of_Reinforcement

Base/Slab_Friction_Coefficient

Crack_Spacing

Mix

Cmnt_Typ

Cmntitious_Cntnt

W/C_Ratio

Ultimate_Shrinkage

Reverse_Shrink

Curing_Type

Strength

Age

Elstc_Modulus

Modulus_of_Rupture

Comp. Strength

Description Load transfer efficiency of the tied concrete shoulder Identifies the presence of a widened lane Width of the widened slab (ft) Level of friction between the base and PCC Base erodobility index Loss of full friction (age in months) Percent steel (%) Bar diameter (in) Steel depth (in) Base/slab friction coefficient Mean crack spacing (in) Mix design properties Cement type Cementitious content Water-cement ratio Ultimate shrinkage Reverse shrinkage Curing type Strength properties Age (yrs) Elastic modulus (psi) Modulus of rupture (psi) Compressive strength (psi)

This table includes a description of each unstabilized/stabilized material data element.

Name

Description

GALTPP_SEC_CON_ID LAYER_NO

Test section identification number. Layer number

Last_Layer (semi-infinite)

Identifies layer as the last layer of the pavement section

Bedrock

Bedrock layer inputs

Type

Soil type

Unit_Weight

Unit weight (pcf)

Poisson_Ratio

Poisson's ratio

B-12

Name Resilient_Modulus Gradation (for each layer) Passing_3_5 Passing_3 Passing_2_5 Passing_2 Passing_1_5 Passing_1 Passing_3/4 Passing_1/2 Passing_3/8 Passing_#4 Passing_#8 Passing_#10 Passing_#16 Passing_#20 Passing_#30 Passing_#40 Passing_#50 Passing_#60 Passing_#80 Passing_#100 Passing_#200 Passing_0_02mm Passing_0_002mm Passing_0_001mm PI LL Compacted_Layer Stabilized Unit_Wght Poisson_Ratio

Resilient modulus (psi)

Description

Gradation inputs for each unstabilized/stabilized layer

Mean percent passing 3- in screen

Mean percent passing 3 in screen

Mean percent passing 2- in screen

Mean percent passing 2 in screen

Mean percent passing 1- in screen

Mean percent passing 1 in screen

Mean percent passing in screen

Mean percent passing in screen

Mean percent passing in screen

Mean percent passing #4 screen

Mean percent passing #8 screen

Mean percent passing #10 screen

Mean percent passing #16 screen

Mean percent passing #20 screen

Mean percent passing #30 screen

Mean percent passing #40 screen

Mean percent passing #50 screen

Mean percent passing #60 screen

Mean percent passing #80 screen

Mean percent passing #100 screen

Mean percent passing #200 screen

Mean percent passing 0.020 mm screen

Mean percent passing 0.002 mm screen

Mean percent passing 0.001 mm screen

Plasticity index

Liquid limit

Compacted layer

Inputs for stabilized layer

Unit weight (pcf

Poisson's ratio

B-13

Name Elastic/Resilient_Mod Minimum_Mod Mod_of_Rupture Therm_Cndctvty Heat_Capacity Strength (for each layer) k1 k2 k3 Poisson_Ratio Ltrl_Pressure Modulus CBR R_Val Lyr_Coefnt DCP

Description Elastic/resilient modulus (psi) Minimum elastic/resilient modulus (psi) Modulus of rupture (psi) Thermal conductivity (BTU/hr-ft-F) Heat capacity (BTU/lb-F) Strength inputs for each unstabilized/stabilized layer Regression constants (used for Level 1 calculation of MR) Regression constants (used for Level 1 calculation of MR) Regression constants (used for Level 1 calculation of MR) Poisson's ratio Lateral pressure Resilient modulus (psi) California Bearing Ratio R-Value AASHTO layer coefficient Dynamic Cone Penetrometer (mm/blow)

B-14

B-15

APPENDIX C: STUDIES RELATED TO SMA
Oklahoma The Oklahoma DOT conducted a study that evaluated and compared the performance of SMA mixes to conventional ODOT S-4 mixes (a Superpave mixture) to determine the performance benefits. Dynamic modulus testing and Hamburg Rut Tests were conducted on S-4 mixes and SMA; performance predicted by the MEPDG was used to evaluate the performance. Hamburg rut depth testing was performed in general accord with OHD L-55. The results showed S-4 mixes had a mean rut depth of 8.41 mm and SMA mixes a mean rut depth of 5.98 mm. SMA mixes had statistically significant lower Hamburg rut depths than S-4 mixes. Dynamic modulus testing was performed at three temperatures and five frequencies in accordance with NCHRP 9-29 PP 02 with the exception of additional test frequencies. The comparisons between SMA and S-4 HMA mixtures at 1 Hz in Figure 26 show that SMA mixes were not as stiff as S-4 HMA mixes at any of the temperatures evaluated. The S-4 mix was 30 to 70 percent stiffer than the SMA mix over the range of temperatures and frequencies tested. SMA, with its lower dynamic modulus, had more total permanent deformation, top-down cracking, bottom-up (alligator) cracking predicted by the MEPDG compared to S-4 mixes. The percent increase in total permanent deformation was not impacted by subgrade resilient modulus and depth to the water table. The lower the subgrade resilient modulus and less depth to water table, the more alligator cracking. Based on the results, the report concluded that it appears that when it comes to asphalt layers, stiffer is better. The MEPDG results seem to go against Hamburg rut test results and published literature on the field performance of the SMA.
C-1

Louisiana A study of Louisiana asphalt mixtures was completed by Mohammad et al., 2007, in which measured E* values of various mixtures were compared with predicted values from the Witczak 1-37A model). The mixtures tested included Superpave mixes designed for high, medium, and low volume roads, SMA mixes, and Marshall mixes. Three types of binder were used: PG 7622M, PG 70-22M and PG 64-22, of which the first two were modified. We looked at three wearing mixes, including I10-2 (12.5 mm Superpave with PG 76-22), I10-3 (12.5 mm SMA with PG 76-22), and I55-2 (12.5 mm Superpave with PG 82-22) to compare the stiffness of SMA and Superpave. Figure 12 shows dynamic modulus at specific temperature and load frequency. The lab measured dynamic modulus for I10-3 (SMA) was slightly lower than the Superpave with same binder (I-10-2). The lower modulus for SMA was consistent in the master curves shown in Figure 20. In addition, dynamic modulus test results obtained from axial and IDT modes showed no statistical differences for the majority of the mixtures tested.
C-2

Maryland. Maryland conducted a study to establish database of material properties for the most common paving materials used in Maryland. The PI found that the Witczak predictive model used for Level 3 dynamic modulus inputs is dominated by temperature influences and does not do a good job of ranking mixtures in terms of their measured stiffness values at a given temperature and loading frequency (Ceylan et al., 2009). In addition, the databases used to develop and calibrate the Witczak and other similar dynamic modulus predictive models contain very few gap-graded
C-3

SMA mixtures of the type commonly used on high volume roads in Maryland. The authors concluded the Witczak predictive equation used to generate the Level 2/3 dynamic modulus data is not intended for SMA mixtures, which is a common premium mixture type in Maryland, and often does not adequately differentiate among different dense graded mixtures. Virginia |E*| tests were performed with the IPC Global (IPC) 100-UTM universal testing machine in accordance with AASHTO TP 62 (AASHTO, 2007a). Five testing temperatures ranging from 14F to 130F and six testing frequencies ranging from 0.1 Hz to 25 Hz were used. The two SMA mixes (08-1025E and 08-1012E) have slightly lower E* compared to the Superpave mixes (08-1036D and 08-1055D).
Studies (e.g., Sotil et al., 2007) have shown that when tested without confinement, certain gapgraded mixtures, such as SMA mixtures, may have lower |E*| values than dense-graded mixtures. SMA mixtures may, therefore, show lower rutting resistance when modeled in the current
C-4

MEPDG software, contrary to the observed superior rutting resistance of SMAs (Michael et al., 2003) in the field. Future studies should, therefore, include confinement to characterize SMA rutting better in the MEPDG when such procedures become standardized.
C-5