<%BANNER%>

Development of specification criteria to mitigate top-down cracking

University of Florida Institutional Repository

PAGE 1

DEVELOPMENT OF SPECIFI CATION CRITERIA TO MITIGATE TOP-DOWN CRACKING By ADAM PAUL JAJLIARDO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Adam Paul Jajliardo

PAGE 3

ACKNOWLEDGMENTS I would like to first thank my advisor and my committee chairman, Dr. Reynaldo Roque for his advice, guidance and support. Without his technical and personal expertise, this would not have been possible. I would also like to acknowledge my other committee members, Dr. Bjorn Birgisson and Dr. Mang Tia, who have lent their knowledge and experience. Special thanks go to Mr. George Lopp for his support in the laboratory and his valuable advice. My deepest thanks go to all the members of the Civil Engineering materials group for their friendship and support during the past two years. They include Tait Karlson, Oscar Garcia, Tipakorn Samarnrak, Jagannatha Katkuri, Claude Villiers, Jeff Frank, JaeSeung Kim, SungHo Kim, Booil Kim, and Boonchi Sangpetngam. I would like to express a very sincere appreciation to my wife Wendy for her love, support, and friendship. I would also like to thank all my family and friends back home who have also supported me during this time. iii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES............................................................................................................vii LIST OF FIGURES...........................................................................................................ix ABSTRACT.............................................................................................................xiii CHAPTER 1 INTRODUCTION........................................................................................................1 1.1 Background.............................................................................................................1 1.2 Objectives...............................................................................................................2 1.3 Scope.......................................................................................................................2 1.4 Research Approach.................................................................................................2 2 LITERATURE REVIEW.............................................................................................4 2.1 Fracture in Asphalt Pavements...............................................................................4 2.2 Mechanisms of Fracture in Asphalt Pavements......................................................4 2.2.1 Traditional Fatigue Approach.......................................................................4 2.2.2 Fracture Mechanics Method.........................................................................6 2.2.3 Dissipated Creep Strain Energy....................................................................7 2.3 Mixture Properties Related to Fatigue Resistance..................................................8 2.3.1 Mixture Stiffness..........................................................................................8 2.3.2 Air Void Content..........................................................................................9 2.3.3 Voids in the Mineral Aggregate (VMA)......................................................9 2.3.4 Asphalt Content and Theoretical Film Thickness........................................9 2.3.5 Binder Viscosity.........................................................................................10 2.3.6 Aggregate Gradation..................................................................................11 2.4 Previous Studies....................................................................................................11 2.5 Summary...............................................................................................................12 3 DESCRIPTION OF TEST SECTIONS......................................................................14 3.1 Locations and Age................................................................................................14 3.2 Pavement Structure...............................................................................................15 3.3 Traffic Volume.....................................................................................................16 iv

PAGE 5

3.4 Environmental Conditions....................................................................................16 3.5 Performance of the Sections.................................................................................16 3.5.1 Overview....................................................................................................16 3.5.2 Field Observations......................................................................................17 4 MATERIALS AND METHODS...............................................................................21 4.1 Extraction of the Field Cores................................................................................21 4.2 Measuring and Cutting the Field Cores................................................................22 4.3 Selecting Samples for Testing..............................................................................22 4.4 Crack Rating.........................................................................................................23 4.5 Mixture Testing....................................................................................................24 4.6 Asphalt Extractions and Binder Testing...............................................................25 4.7 Aggregate Tests....................................................................................................25 4.8 Volumetric Properties...........................................................................................26 5 ANALYSIS AND FINDINGS...................................................................................29 5.1 Volumetric Properties and Extraction-Recovery Results.....................................29 5.11 Air Void Content.........................................................................................29 5.1.2 Effective Asphalt Content..........................................................................31 5.1.3 Aggregate Gradation..................................................................................31 5.1.4 Theoretical Film Thickness........................................................................36 5.1.5 Binder Viscosity.........................................................................................37 5.2 Mixture Results.....................................................................................................37 5.2.1 Resilient Modulus.......................................................................................38 5.2.2 Creep Compliance......................................................................................39 5.2.3 Tensile Strength..........................................................................................40 5.2.4 Failure Strain..............................................................................................40 5.2.5 m-value.......................................................................................................41 5.2.6 Fracture Energy Density and Dissipated Creep Strain Energy..................42 5.3 Non-Destructive Testing (FWD)..........................................................................44 5.3.1 Pavement Structures...................................................................................44 5.3.2 Loading Stresses.........................................................................................46 5.4 Crack Growth Model............................................................................................47 5.5 Individual Analysis of the Sections......................................................................53 5.5.1 I-75 1U and 1C...........................................................................................53 5.5.2 I-75 2U and 3C...........................................................................................54 5.2.3 SR-80 2U and 1C........................................................................................54 6 FURTHER ANALYSIS.............................................................................................56 6.1 Section Data and Mixture Test Results................................................................56 6.2 Mixture Fracture Toughness.................................................................................57 6.3 Mixture Properties................................................................................................64 6.4 Traffic...................................................................................................................71 v

PAGE 6

7 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS.............................76 7.1 Summary and Conclusions...................................................................................76 7.2 Recommendations.................................................................................................77 APPENDIX A SUMMARY OF NON-DESTRUCTIVE TESTING (FWD).....................................78 B SUMMARY OF FDOT FLEXIBLE PAVEMENT CONDITION SURVEY...........94 C SUMMARY OF VOLUMETRIC PROPERTIES......................................................97 D MIXTURE TEST RESULTS...................................................................................100 E SUMMARY OF CRACK GROWTH MODEL RESULTS.....................................105 F SUMMARY OF SECTION DATA FOR ALL SECTIONS....................................109 LIST OF REFERENCES................................................................................................116 BIOGRAPHICAL SKETCH...........................................................................................118 vi

PAGE 7

LIST OF TABLES Table page 3.1 Location of the Sections...........................................................................................14 3.2 Age of the Sections..................................................................................................15 3.3 Thickness of the layers (in)......................................................................................15 3.4 Layer Moduli for each Section (ksi)........................................................................16 3.5 Traffic Volumes for each Section (Millions)...........................................................16 4.1 Average thickness....................................................................................................23 4.2 Average Bulk Specific Gravity................................................................................23 4.3 Cracking Criteria (After Sedwick, 1998).................................................................24 4.4 Crack Ratings...........................................................................................................24 5.1 Air void content........................................................................................................30 5.2 Layer Moduli from FWD Analysis..........................................................................45 5.3 E1/E2 Ratios.............................................................................................................46 A.1 Deflection (mils0 From I-75 1U...............................................................................79 A.2 Deflection (mils) From I-75 1C...............................................................................79 A.3 Deflections for I-75 2U............................................................................................82 A.4 Deflections for I-75 3C............................................................................................82 A.5 Deflections for SR-80 2U.........................................................................................85 A.6 Deflections for SR-80 1C.........................................................................................85 C.1 Bulk Specific Gravity for each Sample....................................................................98 C.2 Effective Asphalt Content, Film Thickness and VMA............................................99 vii

PAGE 8

D.1 Resilient Modulus..................................................................................................101 D.2 Tensile Strength......................................................................................................101 D.3 Creep Compliance..................................................................................................102 D.4 M-value, Failure Strain, Fracture Energy, DCSE, Initial Tangent Modulus, D0, and D1.............................................................................................................103 D.5 Dissipated Creep Strain Energy Calculation..........................................................104 E.1 Estimated Loading Stresses (psi)...........................................................................106 E.2 N f to Initiation for DCSE.......................................................................................107 E.3 N f to Initiation for FE.............................................................................................107 E.4 N f to Propagate 50mm............................................................................................108 F.1 Summary of Traffic Loading (Annual ESALS in 1000)........................................110 F.2 Summary of Loading Stress (psi)...........................................................................110 F.3 Summary of Mixture Test Results Needed for Cracking Model............................111 F.4 Number of Cycles to Propagate Crack Length of 50mm.......................................111 viii

PAGE 9

LIST OF FIGURES Figure page 2.1 Fatigue Crack Growth Behavior (after Jacobs, 1995)................................................7 2.2 Dissipated Creep Strain Energy (after Zhang et al., 2001)........................................8 3.1 Overview of Uncracked I-75 Section.......................................................................18 3.2 Overview of Cracked I-75 Section...........................................................................18 3.3 Overview of SR-80 2U.............................................................................................19 3.4 Overview of SR-80 1C.............................................................................................19 3.5 Longitudinal Crack from SR-80 1C.........................................................................20 4.1 Cutting Machine.......................................................................................................26 4.2 IDT Testing Device..................................................................................................27 4.3 Dehumidifying Chamber..........................................................................................27 4.4 Temperature Controlled Chamber............................................................................28 4.5 Testing Sample with Extensometers Attached.........................................................28 5.1 Air Void Content and Comparison Between WP and BWP Sections......................30 5.2 Effective Asphalt Content (%).................................................................................31 5.3 Gradation Curves for I-75 1U and 1C......................................................................33 5.4 Gradation Curves for I-75 2U and 3C......................................................................34 5.5 Gradation Curves for SR-80 1C and 2U..................................................................35 5.6 Film Thickness (m)................................................................................................36 5.7 Binder Viscosity (Poise)...........................................................................................37 5.8 Resilient Modulus (GPa)..........................................................................................38 ix

PAGE 10

5.9 Creep Compliance at 100 sec. (1/GPa)....................................................................39 5.10 Tensile Strength (MPa)............................................................................................40 5.11 Failure Strain (microstrain)......................................................................................41 5.12 m-value.....................................................................................................................42 5.13 Fracture Energy Density (KJ/m 3 ).............................................................................43 5.14 DCSE (KJ/m 3 )..........................................................................................................44 5.15 Loading stresses (psi)...............................................................................................47 5.16 Number of Cycles to Failure for DCSE at 0 C.......................................................49 5.17 Number of Cycles to Failure for DCSE at 10 C.....................................................49 5.18 Number of Cycles to Failure for DCSE at 20 C.....................................................49 5.19 Number of Cycles to Failure for FE at 0 C.............................................................50 5.20 Number of Cycles to Failure for FE at 10 C...........................................................50 5.21 Number of Cycles to Failure for FE at 20 C...........................................................50 5.22 Comparison between N f of DCSE and FE limits at 0 C.........................................51 5.23 Comparison between N f of DCSE and FE limits at 10 C.......................................51 5.24 Comparison between N f of DCSE and FE limits at 10 C.......................................51 5.25 Number of Cycles to Failure to 50mm at 0 C.........................................................52 5.26 Number of Cycles to Failure to 50mm at 10 C.......................................................52 5.27 Number of Cycles to Failure to 50mm at 20 C.......................................................52 6.1 D 1max for values of DCSE and m-value....................................................................57 6.1 N f to propagate 50mm..............................................................................................58 6.3 Relationship between a, S t, and .............................................................................60 6.3 K HMA Ratio for all sections.......................................................................................62 6.4 K HMA Ratio for sections 0.75 KJ/m 3 < DCSE f < 2.5 KJ/m 3 .....................................63 6.5 Gradations of Low K HMA Ratio Sections with 12.5mm Nominal Aggregate Size..65 x

PAGE 11

6.6 Gradations of High K HMA Ratio Sections with 12.5mm Nominal Aggregate Size..66 6.7 Gradations of Low K HMA Ratio Sections with 9.5mm Nominal Aggregate Size....67 6.8 Gradations of High K HMA Ratio Sections with 9.5mm Nominal Aggregate Size....68 6.9 Case 1 Gradation Comparison for 12.5mm Nominal mixes....................................69 6.10 Case 2 Gradation Comparison for 12.5mm Nominal mixes....................................69 6.11 Case 1 Gradation Comparison for 9.5mm Nominal mixes......................................70 6.12 Case 2 Gradation Comparison for 9.5mm Nominal mixes......................................70 6.13 Relationship Between N f and Required K HMA Ratio................................................71 6.14 FT vs. Traffic loading (ESALS/year x 1000)...........................................................72 6.15 Relationship Between Traffic Level and F S .............................................................73 6.16 F S vs. Traffic (ESALS/year x 1000).........................................................................74 6.17 Minimum K HMA Ratio Required vs. Traffic Levels.................................................75 A.1 Deflections for I-75 1U............................................................................................80 A.2 Deflections for I-75 1C............................................................................................81 A.3 Deflections for I-75 2U............................................................................................83 A.4 Deflections for I-75 3C............................................................................................84 A.5 Deflections for SR-80 2U.........................................................................................86 A.6 Deflections for SR-80 1C.........................................................................................87 A.7 Measured and Computed Deflections for I-75 1U Location 03...............................88 A.8 Measured and Computed Deflections for I-75 1U Location 06...............................88 A.9 Measured and Computed Deflections for I-75 1U Location 09...............................88 A.10 Measured and Computed Deflections for I-75 1C Location 03...............................89 A.11 Measured and Computed Deflections for I-75 1C Location 08...............................89 A.12 Measured and Computed Deflections for I-75 1C Location 10...............................89 A.13 Measured and Computed Deflections for I-75 2U Location 02...............................90 xi

PAGE 12

A.14 Measured and Computed Deflections for I-75 2U Location 09...............................90 A.15 Measured and Computed Deflections for I-75 2U Location 02...............................90 A.16 Measured and Computed Deflections for I-75 1C Location 03...............................91 A.17 Measured and Computed Deflections for I-75 1C Location 04...............................91 A.18 Measured and Computed Deflections for I-75 1C Location 06...............................91 A.19 Measured and Computed Deflections for SR-80 2U Location 05...........................92 A.20 Measured and Computed Deflections for SR-80 2U Location 09...........................92 A.21 Measured and Computed Deflections for SR-80 2U Location 01...........................92 A.22 Measured and Computed Deflections for SR-80 1C Location 02............................93 A.23 Measured and Computed Deflections for SR-80 1C Location 03............................93 A.23 Measured and Computed Deflections for SR-80 1C Location 03............................93 B.1 Cracking Ratings from I-75 1U and 1C...................................................................95 B.2 Cracking Ratings from I-75 2U and 3C...................................................................95 B.3 Cracking Ratings from SR-80 2U and 1C................................................................96 F.1 Effective Asphalt Content (%)...............................................................................112 F.2 Percent Air Voids (%)............................................................................................113 F.3 Theoretical Film Thickness (microns)...................................................................114 F.4 VMA (%)................................................................................................................115 xii

PAGE 13

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering DEVELOPMENT OF SPECIFICATION CRITERIA TO MITIGATE TOP-DOWN CRACKING By Adam Paul Jajliardo May 2003 Chair: Dr. Reynaldo Roque Major Department: Civil and Coastal Engineering One of the most common types of pavement distress in the State of Florida is surface initiated longitudinal wheel path cracking, which is commonly refereed to as top-down cracking. Prior research has indicated that top-down cracking is initiated by critical tensile stresses in the surface of the asphalt pavement due to modern radial truck tires. These cracks then propagate downward by the combined effects of temperature and loads. Several roadway sections were chosen for this study that exhibited top-down cracking. Crack free sections were also studied that had similar structures and traffic as the cracked sections. Asphalt Concrete cores were taken from each section for laboratory testing to determine properties and characteristics of the mixture, binder, and aggregate. xiii

PAGE 14

Traffic volume, structure, and age data were collected for each section and Falling Weight Deflectometer (FWD) tests were performed to determine the moduli of the structural layers. The mixture was tested using the Superpave Indirect Tensile Test (IDT). Resilient modulus, creep compliance, and tensile strength tests were performed. The results of the IDT test were used in the crack growth model developed at the University of Florida to determine whether the number of cycles predicted to initiate and to propagate cracking in the pavement correlated well with observed field cracking performance. The analysis led to the identification of a mixture fracture toughness parameter (K HMA(min) ), which allows for the evaluation of mixture top-down cracking performance by incorporating the affects of mixture properties and pavement structural characteristics. The parameter did an excellent job indicating cracked from uncracked sections. It was concluded that the parameter is suitable for the development of specification criteria to mitigate top-down cracking and specific recommendations were made for its implementation. The parameter (K HMA Ratio) was also defined that allows pavement sections to be compared according to their cracking resistance. xiv

PAGE 15

CHAPTER 1 INTRODUCTION 1.1 Background Surface-initiated longitudinal wheel path cracking is one of the most common types of pavement distress in Florida today. Conventional load-induced cracking has been commonly assumed to initiate at the bottom of the pavement and propagate upwards. Cores and trenched sections taken from substandard sections have clearly illustrated the phenomenon of top down cracking. Myers (1997) identified the potential mechanisms of longitudinal surface initiated wheel path cracking. She determined that this mode of distress is caused by a tensile failure due to high stresses under the ribs of radial truck tires combined with thermal stresses. Myers (2000) also found that surface initiated longitudinal cracking occurs primarily under critical conditions. Therefore, existing design and evaluation methods that consider only average conditions are not adequate in explaining this distress. These approaches do not characterize the actual contact stresses or the discontinuities that exist in the field. Myers (2000) also found that temperature gradients had a strong effect on the development of stresses, which are also not considered in traditional fatigue approaches. Through the study of field cores, Garcia (2002) found that that there was no clear relationship between any one mixture property and the mixture performance. However, he found that the cracking model developed at the University of Florida appeared to adequately explain the differences in mixture performance observed in the field. 1

PAGE 16

2 Surface initiated longitudinal wheel path cracking has resulted in significant rehabilitation costs. A better understanding of the mechanisms of surface cracking and the key mixture properties and characteristics is necessary. The development of specifications and design criteria will result in more crack resistant mixtures. 1.2 Objectives The primary objectives of this research are summarized below: Evaluate field sections to identify the mixture properties and characteristics that most strongly influence surface cracking performance. Develop a design specification for asphalt mixtures that would mitigate surface cracking in pavements. 1.3 Scope This study focuses on the analysis of the key mixture properties and characteristics that affect surface initiated longitudinal wheel path cracking. To accomplish this, over 300 cores were extracted from six field sections throughout the state of Florida. Data was also included from past studies and includes a total of twenty two sections. 1.4 Research Approach The first step in this study was to conduct a literature review in order to understand the different approaches to understanding fatigue in pavements. The different mixture properties and characteristics were also evaluated to determine their influence in the cracking performance of mixtures. Next, the field sections were chosen from a number of potential sections through visual inspection. Data from each of these sections was collected including the age, structure, and traffic. Cracking performance over the life of the pavement was taken from the FDOT Flexible Pavement Condition Survey Database.

PAGE 17

3 Field cores were extracted from the roadway and each core was measured and cut into test specimens. Mixture properties were determined using the Superpave Indirect Tension Test (IDT). The specimens were broken down and the binder was extracted. Binder and aggregate were obtained for evaluation. Falling Weight Deflectometer (FWD) tests were performed on all of the sections in order to determine the moduli of each pavement layer. The results were used to calculate pavement stresses which were used to predict number of cycles to failure using the crack growth model developed at the University of Florida. The results of the crack growth model were analyzed in combination with results from past studies at the University of Florida. The purpose of this was to evaluate the effects of structure, environment, and mixture properties and characteristics on cracking performance to develop specification criteria that would mitigate surface cracking in asphalt pavements.

PAGE 18

CHAPTER 2 LITERATURE REVIEW A literature review was undertaken in order to understand the mixture characteristics and properties that affect crack development and propagation. Several different fatigue approaches were reviewed and their significance was determined when discussing longitudinal surface-initiated top down cracking. It was also important to review previous studies that investigated surface cracking in the field. 2.1 Fracture in Asphalt Pavements Among all the types of failure in pavement, cracking is one of the most predominant. Many factors influence cracking in pavement such as the pavement structure and the mixture characteristics. There are two main types of cracking in asphalt pavements. These are thermal cracking and fatigue cracking. Thermal cracking is caused by the stresses that are induced when low ambient temperatures cool the surface of the road. Fatigue cracking is associated with traffic loading and is generated through repeated stresses. Myers (1997) found that a probable cause of longitudinal surface initiated wheel path cracking is the high tensile stresses caused by modern radial truck tires at the tire-pavement interface. These stresses may be intensified by thermal stresses at the surface. 2.2 Mechanisms of Fracture in Asphalt Pavements 2.2.1 Traditional Fatigue Approach The traditional fatigue approach is based on the assumption that the maximum tensile strains are located at the bottom of the asphalt concrete layer. These strains 4

PAGE 19

5 develop cracks and propagate from the bottom upward into the AC layer. Several fatigue models have been developed to explain this phenomenon. One of the first fatigue models was presented by Monismith et al. (1985). The following relationship defines the fatigue behavior of a particular mixture: cmixbtfSAN11 where, N f is the number of load applications to failure, A is a factor based on asphalt content and degree of compaction, t is the tensile strain, S mix is the mixture stiffness and a and b are constants determined from beam fatigue tests. The Asphalt Institute developed the following empirical relationship in 1982 for a standard mix with an asphalt volume of 11% and an air void volume of 5%: 854.291.3*0796.ENtf where, N f is the number of load applications to cause fatigue cracking in 20% of the pavement area, t is the tensile strain at the bottom of the surface layer, and E* is the dynamic modulus of the asphalt mixture. Another equation used to calculate the fatigue life of a mixture was developed under the SHRP program (Sousa et al., 1996). As in the previous two equations for fatigue life, it is a function of the mixture stiffness and asphalt content. 720.20077.00510738.2SexSNVFBff where, N f is the number of load cycles to failure, e is the base of the natural logarithm, VFB is the voids filled with bitumen, t is the tensile strain, S 0 is the loss of stiffness, and S f is a factor that converts laboratory measurements to anticipated field results. The value of S f is 10 for a pavement that is 10% cracked.

PAGE 20

6 All of these models show that there are many variables that affect the fatigue cracking performance of asphalt mixtures including mixture stiffness, AC content and air voids. Also, this shows that there is no simple or reliable way to predict the fatigue life of an asphalt mixture. Myers (2000) found that the addition of a stiffness gradient in cracked asphalt concrete significantly increased the tensile stresses in the surface of the AC layer. None of the traditional fatigue approaches considers discontinuities (i.e. the presence of a crack) in the asphalt layer or stiffness gradients in the asphalt layer that may be caused by temperature or aging. The position of the load was also found to be a contributing factor. Traditional approaches also do not allow for the possibility of changes in the load positioning (wander) in the field. She concluded that current methods for the design and evaluation are inadequate for longitudinal top-down cracking because they consider only average conditions and this mechanism occurs primarily under critical conditions. 2.2.2 Fracture Mechanics Method Another method to explain fracture in asphalt mixtures is the fracture mechanics method, which introduces the concept of crack propagation. The rate of crack propagation can be predicted using the following relationship known as Paris Law: nKAdNda where a is the crack length, N is the number of load repetitions, A and n are parameters depending on the mixture and K is the difference between maximum and minimum stress intensity factors during repeated loading. According to Ewalds and Wanhill (1986), the fracture mechanics approach identifies three different stages. These are the initiation phase where micro-cracks develop, the propagation phase where the micro

PAGE 21

7 cracks develop into macro-cracks and where crack growth becomes stable, and the disintegration phase where the material fails, and crack growth is unstable. Figure 2.1. Fatigue Crack Growth Behavior (after Jacobs, 1995) 2.2.3 Dissipated Creep Strain Energy Roque et al. (1997) found that the Dissipated Creep Strain Energy (DCSE) limit is one of the most important factors that control crack performance in asphalt concrete mixtures. The DCSE limit is the difference between the fracture energy (FE) and the elastic energy (EE) at the instant of failure. The fracture energy is obtained from a strength test as the area under the stress strain curve up to the point where the specimen begins to fracture. The elastic energy can be obtained from the resilient modulus (MR). Zhang (2000) introduced the concept of a threshold between micro-damage and macro-cracking. Micro-damage was defined to be damage that was determined to be healable. Macro-cracking was determined to be non-healable damage, even over long rest periods and temperature increases. Zhang (2000) found that if the threshold was not reached, cracks would not initiate and the mixture would be able to heal. Conversely, if the threshold was reached the crack would grow and the mixture would not be able to

PAGE 22

8 heal. She determined that the dissipated creep strain energy limit (DCSE f ) was a suitable threshold. Zhang introduced a fundamental crack growth law that is based on this work. Figure 2.2. Dissipated Creep Strain Energy (after Zhang et al., 2001) 2.3 Mixture Properties Related to Fatigue Resistance Many different material properties influence the fatigue resistance of asphalt concrete mixtures. Therefore, it is necessary to review each of these properties to obtain a clear understanding of fatigue resistance in asphalt pavements. 2.3.1 Mixture Stiffness The mixture stiffness is defined as the ratio of the stress to the strain. For asphalt mixtures, the stiffness is a function of time, temperature, and loading. The stiffness of an asphalt mixture is affected by the binder stiffness, gradation, air void content, and asphalt content. As a mixture ages the stiffness increases due to oxidation of the binder. This increases the stiffness of the mixture and produces a mix that is more brittle and less crack resistant.

PAGE 23

9 2.3.2 Air Void Content The amount of permeable air voids in a mix is related to the degree that the binder is exposed to air and water. The exposure of binder to air and water results in the oxidation of the binder and an increase in the rate of age hardening. The increase in age hardening increases the stiffness and brittleness of a mixture. The air void content is a function of aggregate gradation and degree of compaction. Monismith et al. (1985) found that by increasing the air void content excessively resulted in a decreased fatigue life. 2.3.3 Voids in the Mineral Aggregate (VMA) VMA is the volume of the inter-granular void space between the aggregate particles of a compacted pavement mixture. This void space includes the air voids and the asphalt not absorbed into the aggregate. VMA is a function of degree of compaction, aggregate gradation, aggregate shape, and air voids. It is an important factor in the durability of asphalt mixtures. Generally, increased VMA values will increase the durability of a mixture. Excessive VMA with high asphalt content will affect the durability adversely because the high binder content tends to allow the aggregate particles to be pushed apart. In the Superpave Mix Design Procedure (SHRP, 1993), minimum VMA is a function of the nominal maximum aggregate size. Nukunya (2001) questioned the use of the same VMA criteria for both fine and coarse mixes and recommended that requirements should differ for the two. 2.3.4 Asphalt Content and Theoretical Film Thickness Asphalt content is a very important factor in the cracking resistance of a mixture. Asphalt content affects many material properties including air void content and film thickness.

PAGE 24

10 Lower asphalt content has been generally associated with inadequate amounts of asphalt in a mixture. Monismith (1981) found that there is an upper limit to the amount of asphalt that can be incorporated in a mixture, but that this limit should be approached in order to increase the fatigue resistance. Pell and Taylor (1969) found that once the optimum asphalt content is exceeded, there will be a decrease in fatigue resistance. Valkering and Van Gooswilligen (1989) found that an approximate 1% decrease in the binder content was found roughly to halve the traffic-related fatigue life. The theoretical asphalt film thickness is a function of the effective asphalt content and the surface area of the aggregate particles. For any given asphalt content, as the surface area of the aggregate particles increases the theoretical asphalt film thickness decreases. Very thin asphalt films contribute to excessive aging of the binder and in turn, more brittle mixes and decreased cracking resistance. Thicker asphalt films contribute to a more flexible and durable mixture. Kandhal and Chakraborty (1996) suggested a minimum asphalt film thickness to produce durable mixtures. They concluded that an optimum film thickness for HMA, compacted to 4 to 5% air void content, should be higher than 9 to 10 microns. 2.3.5 Binder Viscosity Pell and Taylor (1969) concluded that an increase in binder viscosity resulted in an increase in fatigue resistance. Malan et al. (1989) concluded that higher viscosity asphalts proved to be more crack resistant on lightly trafficked roads, while lower viscosity asphalts resulted in better crack resistant mixtures on highly trafficked roads. This can be explained by the constant kneading effect of the moving loads on high traffic pavements. This kneading effect brings the volatiles to the surface of the pavement and prevents excessive viscosity gradients.

PAGE 25

11 The viscosity of an asphalt binder is influenced by aging and maybe more importantly, by temperature. To prevent premature cracking, the binder viscosity is chosen based on the climate of the region where the mixture will be placed. In low temperature climates, unusually low viscosity binders should not be used because of the risk of extreme temperature shrinkage. 2.3.6 Aggregate Gradation Aggregate gradation plays a very important role in the structure of a mixture. The quality of aggregate interlock is primarily responsible for the mixtures response to load. The aggregate gradation affects VMA and asphalt film thickness. The opinions on the effect of gradation on fatigue resistance are divided. Monismith et al. (1985) found there is an insignificant effect on fatigue resistance that is not explained by air void content and asphalt content. Malan et al. (1989) concluded that continuously graded asphalt mixture designs are less susceptible to surface cracking than gap graded and semi-gap-graded designs. Continuously-graded mixtures tend to have higher asphalt film thickness and are more able to dissipate the shrinkage stresses. 2.4 Previous Studies Sedwick (1998) conducted a study on top-down longitudinal wheel path cracking that examined cores taken from the field. He used these cores to identify the mixture properties and characteristics that would lead to the development of surface cracking. He determined that fracture energy density was a good indicator of cracking performance when other conditions such as pavement structure, traffic, and environmental effects are the same. He also found that samples from the field with fracture energy densities lower that 1.0KJ/m3 indicated a poor crack resistant mixture.

PAGE 26

12 Garcia (2002) also conducted a study on longitudinal wheel path cracking using field cores. He used these cores to identify the key factors that contribute to the development of surface cracking. He determined that there was no clear relationship between any single material property that would adequately describe the cracking performance. However, the results of the HMA fracture mechanics based model developed at The University of Florida appeared to properly explain the difference in mixture performance. He also determined that the effects of the pavement structure and thermal stresses were significant when comparing the relative cracking performance of pavement sections. 2.5 Summary Traditional fatigue approaches do not adequately explain the phenomenon of longitudinal wheel path cracking. Fracture mechanics provides a solid foundation for understanding cracking in asphalt pavements. It introduces the concept of initiation, propagation, and disintegration. Dissipated Creep Strain Energy is one of the most important factors when considering the cracking performance of asphalt mixtures. The Dissipated Creep Strain Energy limit (DCSE f ) can be used as a threshold between micro-damage and macro cracking. Mixture stiffness is a function of temperature, time, and loading. Excessively stiff mixtures are generally less crack resistant. Permeable air voids affect the degree of age hardening. Excessively high air voids will decrease the crack resistance of a material. Film thickness is a function of gradation and asphalt content. Thicker film thickness results in a mixture that is more durable, flexible, and crack resistant. Aggregate gradation plays a defining role in the structure of a mixture. Mixtures that are more continuously graded are more crack resistant. In previous studies, it was found that there was no clear relationship between any one-mixture property and cracking performance. All of the material properties, as

PAGE 27

13 well as the pavement structure must be examined in order to describe a mixtures cracking performance. The crack growth model developed at The University of Florida appears to adequately represent the cracking mechanisms of asphalt mixtures in the field.

PAGE 28

CHAPTER 3 DESCRIPTION OF TEST SECTIONS Six sections from three locations were chosen for this study. These sections were chosen in pairs of good and poor performance with similar structure, loading, and age, but with different mixtures. This chapter provides a description of the sections. 3.1 Locations and Age The six sections were all extracted from locations in southwest Florida. Sections 1U and 1C were taken from I75 in Charlotte County. Sections 2U and 3C were taken from I75 in Lee County. SR 80 sections were also taken from Lee County. Table 3.1 summarizes the locations of the sections. Table 3.1. Location of the Sections Section Number Section Name Condition Code County Section Limits State Mile Posts 1 Interstate 75 U I75-1U Charlotte MP 149.3 MP 161.1 0 11.8 Section 1 2 Interstate 75 C I75-1C Charlotte MP 161.1 MP 171.3 11.8 22.0 Section 1 3 Interstate 75 U I75-2U Lee MP 115.1 MP 131.5 0 16.4 Section 2 4 Interstate 75 C I75-3C Lee MP 131.5 MP 149.3 16.4 34.1 Section 3 5 State Road 80 C SR 80-2C Lee From East of CR 80A 10.8 13.6 Section 1 To West of Hickey Creek Bridge 6 State Road 80 U SR 80-1U Lee From Hickey Creek Bridge 13.6 18.3 Section 2 To East of Joel Blvd. The age of the sections is defined as the time from the most recent resurfacing. The age of each section is summarized in Table 3.2. 14

PAGE 29

15 Table 3.2. Age of the Sections Section Year Let Age as of 2003 I75-1U 1989 14 I75-1C 1988 15 I75-2U 1989 14 I75-3C 1988 15 SR 80-2U 1984 19 SR 80-1C 1987 16 3.2 Pavement Structure The layer moduli were determined with the Falling Weight Deflectometer (FWD). The values were then back calculated using elastic layer analysis. The FWD procedure used the standard SHRP configuration for the sensors (i.e. 8, 12, 18, 24, 36, and 60). For each section, ten tests were conducted in the travel lane in the wheel path at relatively undamaged locations, on both sides of the coring area. A half-inch hole was drilled in the pavement and filled with mineral oil or glycol for heat transfer and the pavement temperatures were recorded. The pavement surface and ambient temperatures were also recorded. A 9-kip seating load was applied, followed by 7, 9, and 11kip loads. Deflection measurements at each of the sensors were recorded. The layer thickness and back-calculated moduli appear in Tables 3.3 and 3.4, respectively. The base and sub-base thickness were not available so a typical thickness of 12 inches was assumed for the back calculation analysis. Table 3.3. Thickness of the layers (in) Section Friction Course AC Base Sub-base I75-1U 0.44 6.23 12 12 I75-1C 0.51 6.54 12 12 I75-2U 0.46 7.42 12 12 I75-3C 0.62 6.47 12 12 SR 80-2U 0.80 6.29 12 12 SR 80-1C 0.37 3.38 12 12

PAGE 30

16 Table 3.4. Layer Moduli for each Section (ksi) Section AC Base Sub-base Sub-grade I75-1U 1000 64 51 36 I75-1C 800 55 50 30 I75-2U 1000 107 90 31 I75-3C 900 60 35 36 SR80-2U 500 57 46 19 SR80-1C 800 44 61 28 3.3 Traffic Volume The traffic volumes for each section are shown in table 3.5. These values are expressed in thousands of ESALS. The traffic volumes vary from 207 K for section SR-80 2U to 674 K for section I-75 3C. Table 3.5. Traffic Volumes for each Section (Millions) Section Traffic (ESALS/year x 1000) I75-1U 558 I75-1C 573 I75-2U 576 I75-3C 674 SR80-2U 207 SR80-1C 221 3.4 Environmental Conditions The environmental conditions were similar for all the test sections. Florida has a humid climate with average yearly temperatures between 20 and 25 C. Pavement temperatures during the summer months can increase considerably. 3.5 Performance of the Sections 3.5.1 Overview The Flexible Pavement Condition Survey Database is a record maintained by the Florida Department of Transportation. This record contains ratings on the ride, rutting, and cracking performance of every pavement section supported by the FDOT. The primary purpose of this record is to prioritize pavement sections for rehabilitation.

PAGE 31

17 The ratings for cracking are based on crack width and cracked surface area. Values between 0 and 10 are given to a pavement section depending on the size of the crack widths and the surface area of the pavement that is cracked. The value of 10 is given to a pavement that is crack free. Appendix B contains the crack ratings of all the sections. Unfortunately, this rating judges only the appearance of the surface and gives no indication of the actual depth of the cracks. For example, a pavement with a high amount of cracking in the friction course would be given a low crack rating even if the cracks did not propagate further into the pavement. Also, a pavement with a small number of cracks would be given a high rating even though the cracks may extend well into the pavement. Therefore, to gauge the actual extent of the cracking, it was necessary to core the pavement directly though the crack and measure the crack depths manually. 3.5.2 Field Observations Before the coring was performed, a field trip was taken to each section to observe and take pictures. Figure 3.1 shows a typical uncracked section for I-75. Figure 3.2 shows a cracked section. The uncracked section appears to be in an acceptable condition. The cracked section exhibits a moderate amount of cracking as well as wheel rim markings. The cracks appear in and to the side of the wheel paths in the travel lane. Figures 3.3 and 3.4 show the uncracked (SR80 2U) and cracked (SR80 1C) sections respectively. The uncracked section appears to be in a very acceptable condition with a surface free from cracks. The cracked section is heavily cracked with continuous cracks appearing in the wheel paths of both lanes. Figure 3.5 shows a close-up of a crack.

PAGE 32

18 Figure 3.1. Overview of Uncracked I-75 Section Figure 3.2. Overview of Cracked I-75 Section

PAGE 33

19 Figure 3.3. Overview of SR-80 2U Figure 3.4. Overview of SR-80 1C

PAGE 34

20 Figure 3.5. Longitudinal Crack from SR-80 1C

PAGE 35

CHAPTER 4 MATERIALS AND METHODS After the cores were extracted from the roadway, they were measured and saw-cut into individual testing specimens. The bulk specific gravity of each specimen was measured. The specimens were then tested using the Superpave indirect tensile test (IDT) developed by Roque et al. (1997). One specimen was used to determine the Maximum Theoretical Density using the Rice test. The binder and the aggregate were separated from representative specimens from each section. These were used for further testing. 4.1 Extraction of the Field Cores Several cores were taken from each of the six sections. Cores were extracted from the wheel path as well as between the wheel paths. Cores were also taken through the crack in order to measure the actual crack depths. The cores were marked for traffic direction. This is necessary because the failure that was observed in the field is a tensile failure perpendicular to the direction of traffic. It was important to keep the direction consistent when performing the IDT tests. A total of 46 cores were extracted from each section. Eighteen cores each were taken from the wheel path and eighteen from between the wheel paths. Ten cores were taken through the cracks for each section. All of the cores were extracted using a truck mounted coring rig. The truck-mounted rig was used to minimize damage that may occur to the samples during the coring process. 21

PAGE 36

22 4.2 Measuring and Cutting the Field Cores Upon inspection in the laboratory, the thickness of each lift was measured and recorded. The crack depths were also measured and recorded. Since the cracks originate at the surface, the layer immediately beneath the friction surface is primarily responsible for crack initiation and propagation. This layer was chosen for the purposes of this study. This layer was identified for each core and marked for cutting. Figure 4.1 shows a picture of the machine used to cut the samples. The thickness of the sample used for the IDT testing is typically between 1 to 2 inches. The actual thickness of the specimens varied from 1 to 1.81 inches, depending on the thickness of the layer as well as the quality of the cores. The average thickness of the samples for each section is shown in Table 4.1. After the specimens were cut, they were marked for future identification. Since the cutting process involves water, the specimens were placed in an air-conditioned environment for several days until their natural moisture content was reached. The bulk specific gravity of each specimen was measured. 4.3 Selecting Samples for Testing Nine specemins were needed from each section in order to test the mixture at three temperatures. The samples were chosen by selecting nine samples having a bulk specific gravity (G mb ) closest to the average G mb of the section. Table 4.2 shows the average bulk specific gravities for each section.

PAGE 37

23 Table 4.1. Average thickness Section Thickness (in.) I75-1U 1.17 I75-1C 1.10 I75-2U 1.00 I75-3C 1.06 SR 80-2U 1.31 SR 80-1C 1.81 Table 4.2. Average Bulk Specific Gravity Average G mb Section WP BWP I75-1U 2.302 2.241 I75-1C 1.860 2.274 I75-2U 2.284 2.221 I75-3C 2.281 2.208 SR 80-2U 2.232 2.181 SR 80-1C 2.267 2.204 Note: WP: Wheel Path BWP: Between Wheel Path 4.4 Crack Rating The cracked cores were taken and the crack depths were measured and recorded. Sedwick (1998) defined a crack rating criteria based on the average crack depth measured for a given section. This criteria assigns a value between 0 and 10 based on the length of the measured crack depths. Table 4.3 shows the rating criteria used by Sedwick. Table 4.4 shows the average crack depths for the six sections and their corresponding performance rating. Cracking was found to be especially severe in section SR-80 1C where in some cases the cracks extended completely through the cores.

PAGE 38

24 Table 4.3. Cracking Criteria (After Sedwick, 1998) Crack Depth (in) Rating < 0.25 10 0.26 0.75 8 0.76 1.25 6 1.26 2.00 4 2.01 3.00 2 > 3.00 0 Table 4.4. Crack Ratings Average Cracking Performance Section Depth (in) Rating I75-1U Uncracked 10 I75-1C 2.21 2 I75-2U Uncracked 10 I75-3C 2.32 2 SR 80-2U Uncracked 10 SR 80-1C 2.56* 2 *Some samples cracked completely through 4.5 Mixture Testing The testing procedures used in this study were developed for the FDOT by Roque et al, (1997). The following tests were performed: Resilient Modulus, Creep Compliance, and Tensile Strength. These tests were performed at three temperatures: 0C, 10C, and 20C. The results from the three tests were analyzed using software developed at the University of Florida. This provided Resilient Modulus (GPa), Creep compliance as a function of time (1/GPa), Tensile Strength (MPa), Failure strain (microstrain), Fracture Energy (KJ/m 3 ), m-value (the slope of the linear portion of the creep compliance-time curve), and the Dissipated Creep Strain Energy to failure. Poissons ratio is also calculated for each of the three tests. A gage placement jig was used to mount four aluminum gage points on each face of the testing specimens. The samples were then placed in a relatively low humidity

PAGE 39

25 chamber for forty-eight hours to eliminate any moisture that would affect testing. Before testing, the samples were placed in the temperature-controlled chamber overnight to stabilize the temperature. Before testing, the samples were fitted with knife edged gage mounting blocks. A set of spring-loaded extensometers was placed on the knife-edges to measure deformations. Figures 4.2 through 4.5 show pictures of the IDT testing machine, the dehumidifying chamber, the temperature-controlled chamber, and a sample with the extensometers attached. 4.6 Asphalt Extractions and Binder Testing Two samples from each section were broken down to determine the Theoretical Maximum Specific Gravity (G mm ) or Rice Gravity according to AASHTO T 209-94. The G mm value for each section made it possible to calculate the air void content for each sample. These samples were then placed in an asphalt extraction device that uses trichloroethylene (TCE) to separate the binder from the aggregate. The binder-TCE mixture was then placed in an extraction device, which evaporated the TCE. Viscosity tests were performed on the binder at 60C using the Brookfield Thermosel Apparatus. These tests were performed according to ASTM D 4402-87. 4.7 Aggregate Tests Once the aggregate was separated from the binder, it was placed in an oven overnight to evaporate any remaining TCE. Following this, a washed sieve analysis was performed according to ASTM C117 to determine the amount of material passing the No. 200 sieve. A dry sieve analysis was then performed to determine the gradation of each mixture following ASTM C136. The specific gravity and absorption of the fine and coarse aggregates was performed according to ASTM C128 and ASTM C127 respectively.

PAGE 40

26 4.8 Volumetric Properties Using the results of the binder and aggregate tests, several volumetric properties were calculated. These were the effective asphalt content, the Voids in the Mineral Aggregate (VMA), and the theoretical film thickness. The effective asphalt content was calculated using the results from the extraction-recovery process as well as the percent absorption. The VMA was calculated using the bulk specific gravity of the mixture, the specific gravity of the aggregate and the aggregate content. The theoretical film thickness was obtained from the Hveem method. The Hveem method calculates the film thickness by approximating the surface of the aggregate using surface area factors. These surface area factors are multiplied by the percentage passing for each sieve. The film thickness is calculated by dividing this surface area by the volume of effective asphalt. The calculations of these volumetric properties are summarized in Appendix C. Figure 4.1. Cutting Machine

PAGE 41

27 Figure 4.2. IDT Testing Device Figure 4.3. Dehumidifying Chamber

PAGE 42

28 Figure 4.4. Temperature Controlled Chamber Figure 4.5. Testing Sample with Extensometers Attached

PAGE 43

CHAPTER 5 ANALYSIS AND FINDINGS This chapter presents the results of the mixture and binder results as well as the analysis of the Falling Weight Deflectometer Test (FWD). This chapter also provides analysis as to the material properties that were related to the cracking performance of the mixtures. This analysis was conducted through the comparison of sections with similar age, pavement structure, and traffic conditions. 5.1 Volumetric Properties and Extraction-Recovery Results The cores that were obtained in the field were cut and specimens were used to determine the bulk specific gravity of the mixture. Samples from each section were broken down and their Theoretical Maximum Specific Gravity was determined. These samples were also put through an extraction and recovery process to determine the asphalt content and viscosity as well as several aggregate properties. From these results, several volumetric properties were calculated including effective asphalt content, VMA, and theoretical film thickness. Results are shown in the sections that follow. 5.11 Air Void Content The air void content for each section was calculated using the bulk specific gravity of the mixture (G sb ) and the theoretical maximum specific gravity (G mm ). The average void content and standard deviation for each of the sections is shown in Table 5.1. The comparisons between the sections are shown in Figure 5.1. 29

PAGE 44

30 Table 5.1. Air void content WP BWP Air Void Standard Air Void Standard Section Content Deviation Content Deviation I75-1U 1.86 0.13 3.21 0.35 I75-1C 2.88 0.18 5.39 0.52 I75-2U 3.72 0.57 6.93 0.82 I75-3C 4.37 0.51 7.24 0.88 SR 80-2U 4.72 1.10 7.53 0.69 SR 80-1C 2.77 0.55 5.71 0.53 0.001.002.003.004.005.006.007.008.00I75-1UI75-1CI75-2UI75-3CSR 80-2U SR 80-1C Section % Air Void Content WP BWP Figure 5.1. Air Void Content and Comparison Between WP and BWP Sections. For all the sections, there seemed to be a similar difference in air void content between the WP cores and the BWP cores. The I-75 sections 1U and 1C showed a difference in air voids with values ranging between 2% and 3% for the WP cores and 3% and 5.5% for the BWP cores. The I-75 sections 2U and 3C exhibited approximately the same percentage of air voids with approximately 4% for the WP cores and 7% for the BWP cores. The SR-80 sections 2U and 1C also showed a difference in air voids. The values ranged from approximately 5% for section 2U to 3% for section 1C for the WP

PAGE 45

31 cores and from 7.5% for 2U and 5% for 1C for the BWP cores. There was no clear relationship between the air void content and the cracking performance. 5.1.2 Effective Asphalt Content The effective asphalt content was determined from the percent asphalt absorbed using the aggregates from the extraction-recovery process. Low asphalt content means poor coating of the aggregate particles and poor fatigue resistance. The FDOT requires effective asphalt contents to be greater than 5%. The figure below shows that none of these sections met the requirement. There was little difference between the cracked and uncracked pairs and no visible relationship between effective asphalt content and cracking performance. 0.001.002.003.004.005.006.00I75-1UI75-1CI75-2UI75-3CSR-80 2USR-80 1CSectionEffective Asphalt Content (%) Figure 5.2 Effective Asphalt Content (%) 5.1.3 Aggregate Gradation The aggregate gradation for each of the sections is shown in figures 5.3 to 5.5. The gradations shown below are expected to be finer than the in place gradations. This is because the aggregates used to compute the distributions were taken from samples cut

PAGE 46

32 from cores and the sizes of the specimens were less than the 2500 grams required for a gradation test. The resulting gradations will be finer than the actual gradations in the field but since the size of the specimens were approximately the same it was assumed that the shift in gradation due to these two effects was the same for all samples. The aggregates from I-75 sections 1U and 1C have a similar aggregate gradation distribution. Each section has low dust contents and a significant amount of material between the #100 and #8 sieves. I-75 sections 2U and 3C are compared in Figure 5.3. The grain size distributions for the two sections are also similar. Each has a low dust content, a significant amount of material between the #100 and #30 sieves and a gap between the #30 and #8 sieves. Section 3C appears to be a slightly more gap-graded mixture. Previous studies (Sedwick, 1998) found that gap graded mixtures are generally less crack resistant than more continuously-graded mixtures. Section 3C appears to be slightly finer than section 2U. SR-80 sections 2U and 1C are compared in Figure 5.4. Their distribution curves are similar in that they both have a large amount of material between the #200 and #50 sieves and both have low dust contents. Section 2U has a finer distribution than section 1C. A finer gradation has also been linked to poor cracking performance because of the greater tendency for the mixtures to have low asphalt film thickness. However, there does not appear to be a strong relationship between the relative fineness of the mixture and the cracking performance.

PAGE 47

0102030405060708090100Sieve sizes% Passing I75-1C I75-1U 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine 33 Figure 5.3 Gradation Curves for I-75 1U and 1C

PAGE 48

0102030405060708090100Sieve sizes% Passing I75-3C I75-2U 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine 34 Figure 5.4 Gradation Curves for I-75 2U and 3C

PAGE 49

0102030405060708090100Sieve sizes% Passing SR80-1C SR80-2U 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine 35 Figure 5.5 Gradation Curves for SR-80 1C and 2U

PAGE 50

36 5.1.4 Theoretical Film Thickness The theoretical film thickness was calculated using the Hveem method. The Hveem method calculates the film thickness from the aggregate gradation and the asphalt content. As previously discussed in Chapter 2, Kandhal proposed a minimum film thickness of 9 to 10m. From the data below in Figure 5.6, all of the sections have theoretical film thickness of less than 9m and are generally below 6m, which is considered excessively low. This may be due to the excessively low effective asphalt contents. For the I-75 1U and 1C and the SR-80 sections the uncracked sections had higher film thickness than the cracked sections, while the reverse was found for the I75 2U and 3C sections. 4.404.604.805.005.205.405.605.80I75-1UI75-1CI75-2UI75-3CSR-80 2USR-80 1CSectionFilm Thickness (m) Figure 5.6 Film Thickness (m)

PAGE 51

37 0100002000030000400005000060000700008000090000100000I75-1UI75-1C I75-2U I75-3C SR 80-2U SR 80-1C SectionBinder Viscosity (Poise) WP BWP Figure 5.7 Binder Viscosity (Poise) 5.1.5 Binder Viscosity Figure 5.7 shows a comparison of the binder viscosities between the sections. The binder viscosity values follow a similar trend with the air void content. Sections with higher air void contents also exhibit higher binder viscosities. This can be explained by the age hardening that occurs due to oxidation. A higher air void content generally allows for an increased amount of oxidation that results in a higher rate of age hardening. The SR80 sections show much higher binder viscosities than other sections. This may be partially due to the older age of the pavement sections. 5.2 Mixture Results The mixture test performed were resilient modulus, creep compliance at 100 seconds and tensile strength. These tests were performed at 0 C, 10 C, and 20 C. The mixture properties that were obtained from these tests were the resilient modulus, creep compliance, m-value, tensile strength, fracture energy density, failure strain, initial

PAGE 52

38 tangent modulus, and the dissipated creep strain energy limit (DCSE f ). The following is a summary of the test results and a detailed analysis of each mixture property and how it relates to mixture cracking performance. 5.2.1 Resilient Modulus The resilient modulus (M R ) is a measure of a materials elastic stiffness. This is a function of the binder stiffness and the degree of aggregate interlock. Figure 5.8 shows the values of M R for each of the sections at each of the three test temperatures. 02468101214161820I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionResilient Modulus (GPa) 0 10 20 Figure 5.8 Resilient Modulus (GPa) The I-75 1U and 1C sections exhibited almost the same M R values for 10 and 20. Section 1C had slightly higher M R values at 0 with values only 13% over those of Section 1U. I-75 sections 2U and 3C also had similar values of M R At 20, the values were almost equal, at 10 section 3C was greater than 2U by 12.5% and at 0 section 2U was greater than 3C by 10%. The results for the SR80 sections are similar to those of the I-75 sections. The resilient modulus values for SR-80 2U and 1C are also close to equal.

PAGE 53

39 At 0 and 10, the results for the two sections are almost identical. At 20 the value for section 2U is higher than 1C by 20%. The data suggests that there is no clear relationship between resilient modulus and cracking performance. However, tensile stresses will be greater in the sections that have slightly greater resilient modulus values. 5.2.2 Creep Compliance Creep compliance is related to the ability of a mixture to relax stresses especially thermal stresses. Mixtures with higher creep compliances can relax stresses more quickly than mixes with low creep compliances. Figure 5.9 shows the creep compliances at 100 seconds for each test section. For the I-75 1U and 1C sections, the compliance values were slightly higher for section 1U than those for 1C. In contrast, the compliance values for I-75 2U and 3C were very similar. For sections SR 80 2U and 1C, the uncracked section had lower compliance values than the cracked section. There appears to be no relationship between creep compliance and the cracking performance of the sections. 0.0000.5001.0001.5002.0002.5003.0003.500I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionCreep Compliance at 100 sec (1/GPa) 0 10 20 Figure 5.9 Creep Compliance at 100 sec. (1/GPa)

PAGE 54

40 5.2.3 Tensile Strength Tensile strength is the maximum tensile stress that the mixture can withstand before failure. The indirect tensile strengths for each section at all three temperatures are shown in Figure 5.10. The uncracked sections possessed slightly higher tensile strengths than the cracked sections. 0.000.501.001.502.002.503.003.50I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionTensile Strength (MPa) 0 10 20 Figure 5.10 Tensile Strength (MPa) 5.2.4 Failure Strain Failure strain is the horizontal strain that is measured during the indirect tensile strength test when cracking occurs. Failure strain is a direct measurement of the brittleness of a mixture. In general, mixtures with high failure strains are more crack resistant. The failure strains for each section at each temperature are shown in Figure 5.11. For all sections, the uncracked sections displayed higher failure strains than the cracked sections excluding I-75 sections 2U and 3C at 20 C, which were close to equal.

PAGE 55

41 050010001500200025003000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionFailure Strain (microstrai n 0 10 20 Figure 5.11 Failure Strain (microstrain) 5.2.5 m-value The m-value is defined as the slope of the linear portion of the log creep compliance log time curve. It is calculated by fitting the following relationship to the creep compliance data: mtDDtD10 where, D(t) is compliance at time t, D 0 and D 1 are model parameters, and m is the m-value. The m-value is an indirect measurement of the creep rate of a mixture. A mixture with a higher m-value has a higher creep rate, which implies a higher rate of damage for a given stress. However, it also means a higher rate of stress relaxation. Also higher m-values are typically associated with softer binders and mixtures with higher Fracture Energy thresholds. Figure 5.12 shows the m-values for each section at all three temperatures. The most significant difference in m-values between paired sections was found on SR-80.

PAGE 56

42 The m-value for section SR-80 1C was higher than section 2U for all temperatures. These values agree with the lower binder viscosity for SR-80 1C. 0.00.10.20.30.40.50.6I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionm-Value 0 10 20 Figure 5.12 m-value 5.2.6 Fracture Energy Density and Dissipated Creep Strain Energy Fracture energy density is defined as the energy per unit volume that is required to fracture an asphalt mixture. It is calculated from the indirect tensile strength test by computing the area under the stress-strain curve up to the point the sample starts to fail. Previous studies (Sedwick, 1997) have determined that fracture energy is a reliable indicator of the crack resistance of a mixture when other conditions such as pavement structure and traffic are similar. He suggested that mixtures with fracture energy densities of less that 1 KJ/m3 at 0 or 10 performed poorly in the field. Garcia (2002) found that the pavement structure and thermal stresses were also important when comparing the relative performance of asphalt mixtures. The fracture energy densities are shown below in Figure 5.13 for all the sections at the three temperatures. As the data indicates, the fracture energy densities for the

PAGE 57

43 uncracked sections are greater than their paired cracked sections with the exception of the I-75 2U and 3C sections at 20. The values of fracture energy for all cracked sections were below 1 KJ/m 3 All uncracked sections had fracture energy values equal to or greater than 1 KJ/m 3 at 10C. The fracture energy of the SR-80 uncracked section at 0C was less than 1 KJ/m 3 The much smaller traffic levels experienced by this section may explain its good performance in the field. 0.00.51.01.52.02.53.0I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionFracture Energy (KJ/m3) 0 10 20 Figure 5.13 Fracture Energy Density (KJ/m 3 ) The dissipated creep strain energy at failure (DCSE f ) is defined as the fracture energy minus the elastic energy (Zhang, 2000). The values of DCSE f for each section are shown below in Figure 5.14 at the three test temperatures. Since DCSE is a function of the fracture energy, it is reasonable that the results would display a similar trend. From the results of the data, it can be seen that the uncracked sections posses higher DCSE values than their respective cracked pairs except for sections I-75 2U and 3C at 20.

PAGE 58

44 0.00.51.01.52.02.53.0I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionDCSE (KJ/m3) 0 10 20 Figure 5.14 DCSE (KJ/m 3 ) 5.3 Non-Destructive Testing (FWD) Falling Weight Deflectometer (FWD) testing was performed on each of the section in order to determine the layer moduli in the pavement system. Back calculation analysis was used to interpret the testing data. These modulus values were used including information about the layer thickness to calculate the stresses at the bottom of the AC layer. The measured deflections for each sensor and location are shown in Appendix A. 5.3.1 Pavement Structures FWD testing was run at three deflection levels: high, intermediate, and low. The deflection basin data was used to back calculate the moduli of each layer in the pavement system using BISDEF at three locations along the length of each section. Table 5.2 shows a summary of the results.

PAGE 59

45 Table 5.2 Layer Moduli from FWD Analysis Section AC Base Sub-base Sub-grade I75-1U 1000 64 51 36 I75-1C 800 55 50 30 I75-2U 1000 107 90 31 I75-3C 900 60 35 36 SR80-2U 500 57 46 19 SR80-1C 800 44 61 28 The results of the back calculation show that there are some differences in the structure of the pavement sections. I-75 sections 1U and 1C have similar base and sub-base moduli although the base modulus for 1U is slightly higher than 1C. One of the largest differences is in the base and sub-base moduli for I-75 sections 2U and 3C. The base and sub-base moduli for section 2U are almost double those of section 3C. This had a significant impact upon lowering the tensile stresses in section 2U. The structures for the SR-80 sections were similar although the base stiffness for section 2U was slightly greater that that of 1C. The sub-base stiffness is higher for section 1C but this may be attributed to fitting error in the back calculation analysis and does not have a significant effect on the pavement stresses. The ratio of the asphalt concrete modulus to the base modulus (E1/E2) is a good indicator of the bending stresses in the AC layer. A larger E1/E2 ratio generally indicates higher bending stresses. Table 5.3 shows the E1/E2 ratios for each section. For each paired cracked and uncracked section, the E1/E2 ratios were larger for the cracked section. For the I-75 sections 2U and 3C, the ratios were doubled for section 3C.

PAGE 60

46 Table 5.3 E1/E2 Ratios Section Temp E1/E2 0 34 I75-1U 10 25 20 17 0 45 I75-1C 10 29 20 21 0 25 I75-2U 10 14 20 10 0 41 I75-3C 10 28 20 20 0 46 SR80-2U 10 34 20 28 0 62 SR80-1C 10 45 20 30 5.3.2 Loading Stresses Using the layer thickness and the layer moduli calculated from the FWD data, the loading stresses were calculated at the bottom of the AC layer with BISAR. These stresses were calculated at three loading levels: 7,000 lbs., 9,000 lbs., and 11,000 lbs. and at the three testing temperatures of 0, 10, and 20. A complete summary of the loading stresses appears in Appendix E. The modulus values used in the stress calculation were the M R values calculated in the laboratory. Figure 5.4 shows the stresses for a 9,000-lb load at all three temperatures. For any given temperature and load, the estimated stresses in the bottom of the AC layer were greater for the cracked sections than for the uncracked sections. The stresses for I-75 sections 1U and 1C are almost identical and indicate similar structural characteristics. The other sections displayed much greater differences the their stresses.

PAGE 61

47 I-75 section 3C had approximately 60% greater stresses than section 2U. The stresses for SR-80 section 1C were in some cases double those of section 2U. 050100150200250300350400450I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionStress (psi) 0 10 20 Figure 5.15. Loading stresses (psi) 5.4 Crack Growth Model As mentioned earlier in Chapter 2, Zhang (2001) developed a fracture mechanics based crack growth model at The University of Florida. This model was used to determine the relative cracking performance of each roadway section. The crack growth model explains two phases of crack development: initiation and propagation. The initiation phase predicts the number of cycles of loading required to exceed a threshold point and begin macro cracks. The propagation phase predicts the number of cycles that are required to grow a crack a given distance starting with an initial user-defined crack length. The stresses that were used as inputs for both phases of the cracking model were calculated using BISAR at the bottom of the asphalt layer. Since the stresses were calculated at the bottom and not at the top of the asphalt layer, the results of the initiation and propagation phases of the cracking model can only be used to compare the cracking performance in a relative manner.

PAGE 62

48 For the initiation phase, the number of cycles required to achieve macro cracking -were calculated for low, medium, and high loading levels (7000, 9000, and 11000 Lbs.) at the three test temperatures of 0, 10, and 20 C. The model calculates the number of cycles for both dissipated creep strain energy and fracture energy limits. The results of the initiation phase appear in Figures 5.16 through 5.21. For each temperature and for each load, it was found that the uncracked sections required more cycles to reach their thresholds and initiate a crack than their paired uncracked sections. In all cases, there were significant differences in N f I-75 section 1U required almost two times the N f of section 1C to reach its threshold. I-75 section 2U required three times the N f of section 3C to reach its threshold and SR-80 section 2U required almost six times the N f of SR-80 section 1C. A comparison between N f for the DCSE limit and N f for the FE limit for all three temperatures with a 9000 lb. load is shown in Figures 5.22 through 5.24. For all cases, N f was lower for DSCE than for FE. This indicates that DSCE is more critical for all of these sections. The results of the propagation phase were also analyzed for these sections individually and appear below in Figures 5.25 through 5.27. The N f for the propagation phase is defined as the number of cycles that are required to propagate the crack from an initial length of 4mm to a final length of 50mm.

PAGE 63

49 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation 7000 9000 11000 Figure 5.16 Number of Cycles to Failure for DCSE at 0 C 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation 7000 9000 11000 Figure 5.17 Number of Cycles to Failure for DCSE at 10 C 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation 7000 9000 11000 Figure 5.18 Number of Cycles to Failure for DCSE at 20 C

PAGE 64

50 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation 7000 9000 11000 Figure 5.19. Number of Cycles to Failure for FE at 0 C 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation 7000 9000 11000 Figure 5.20. Number of Cycles to Failure for FE at 10 C 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation 7000 9000 11000 Figure 5.21. Number of Cycles to Failure for FE at 20 C

PAGE 65

51 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation DCSE FE Figure 5.22. Comparison between N f of DCSE and FE limits at 0 C 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation DCSE FE Figure 5.23. Comparison between N f of DCSE and FE limits at 10 C 01000020000300004000050000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Initiation DCSE FE Figure 5.24. Comparison between N f of DCSE and FE limits at 10 C

PAGE 66

52 0500010000150002000025000300003500040000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Propogate 50mm 7000 9000 11000 Figure 5.25. Number of Cycles to Failure to 50mm at 0 C 0500010000150002000025000300003500040000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Propogate 50mm 7000 9000 11000 Figure 5.26. Number of Cycles to Failure to 50mm at 10 C 0500010000150002000025000300003500040000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSectionNf to Propogate 50mm 7000 9000 11000 Figure 5.27. Number of Cycles to Failure to 50mm at 20 C

PAGE 67

53 For each case at all temperatures and loadings, the uncracked sections required more cycles to propagate a crack to 50mm than their paired cracked sections. There were significant differences in N f between the paired sections. I-75 section 1U required approximately twice the cycles of section 1C to propagate to 50mm. I-75 section 2U required close to 2.5 times the cycles of section 3C. SR-80 section 2U required over six times the cycles of section 1C. The values of N f for SR-80 section 1C are the same across all three temperatures due to the extremely high stress that this section experienced. This caused the material to fail immediately according to the cracking model. 5.5 Individual Analysis of the Sections 5.5.1 I-75 1U and 1C Low dissipated creep strain energy was the primary reason for the failure of section I-75 1C. 1C and 1U had similar structures, m-value and D 1 The DCSE value of 1U was almost two times that of section 1C. From the extraction-recovery results it was found that 1U had a much lower air void content, slightly lower asphalt content but higher film thickness than section 1C. The air void content of section 1C was 68% higher than section 1U. From the gradation curves, there was no significant difference between the two sections. Both were very similar and followed the same trend. The binder viscosity test showed that 1C had 33% higher viscosity than 1U. The mixture property tests showed that the M R and the creep compliance values were close to equal. Section 1C had almost 15% higher tensile strength than 1U and the failure strain for section 1U was almost twice that of section 1C.

PAGE 68

54 5.5.2 I-75 2U and 3C The probable causes for failure in section 3C are related to the structural effects and low DCSE values. The stresses calculated at the bottom of the asphalt for 3C are 60% higher than those for 2U. These higher stress values resulted from lower modulus base and sub base layers. The extraction-recovery results show significant differences. 3C had slightly higher AC contents and 5% greater film thickness. The gradation results showed that while both mixtures tended to be gap graded, section 3C was slightly more gap graded. The binder viscosity for section 3C was approximately 25% greater than for section 2U. The mixture test results also displayed important differences between the two sections. While the M R values were similar, the creep compliance results at 100 sec were slightly lower for 3C at 10 C. The tensile strength of 3C was 25% lower than that of 2U. Also, failure strain for 2U was 45% higher than for 3C. 5.2.3 SR-80 2U and 1C The cause for failure in section 1C was due to structural effects, low DCSE values, and high m-values. The calculated stresses for section 1C were extremely high and were 50% higher than those for section 2U. This was at least partially due to the low thickness of the AC layer. The trend in air void content for these two sections is the opposite of the I-75 sections. Section 2U has as much as 70% higher air voids than section 1C. The results of the extraction-recovery process also show many differences. Section 2U had slightly higher AC content and film thickness than 1C. The viscosity of the binder was 86% greater for the uncracked section. Although the gradation plots showed a similar trend, 2U was a much finer mix than section 1C.

PAGE 69

55 The mixture test results also showed significant differences. The MR results were close to equal for the two sections. The creep compliance however, for section 1C was 90% greater than that of section 2U. The ultimate tensile strength of 2U was 20% higher than 1C and the failure strain was slightly higher. The m-value and DCSE values were also significant. For all temperatures, the cracked section had much higher m-values and section 2U had almost 40% greater DCSE values.

PAGE 70

CHAPTER 6 FURTHER ANALYSIS This chapter presents the results of an analysis of the pavement sections previously discussed as well as pavement sections studied by Garcia (2002) and Sedwick (1998). This chapter also introduces two parameters: the minimum mixture fracture toughness as well as the mixture fracture toughness ratio. 6.1 Section Data and Mixture Test Results A summary of the pavement section data for the sections studied by Garcia and Sedwick are included in Appendix F. These include traffic and the stresses calculated from BISAR. The traffic is displayed in units of millions of ESAL S /Year. The stresses were calculated as the maximum tensile stress at the bottom of the asphalt layer. A thorough presentation of the additional section data including layer thickness and moduli appears in Garcia (2002) and Sedwick (1998). A summary of the mixture test results for all sections also appears in Appendix F. As mentioned earlier, the values for m, D 0 and D 1 are calculated from fitting the following equation to the creep compliance test results: mtDDtD10 where, D(t) is compliance at time t, D 0 and D 1 are model parameters, and m is the m-value. It should be noted that the D 1 and m-value parameters were calculated by using a constant D 0 value of 3.3310 -7 psi for all sections. This was done to provide consistent values of D 1 and m. 56

PAGE 71

57 6.2 Mixture Fracture Toughness The number of cycles to propagate a crack 50 mm was calculated with the HMA crack growth model for each section. Figure 6.1 shows the comparisons of N f Each uncracked section clearly requires more cycles to achieve a crack length of 50mm than its paired cracked section. It should be noted that SR-16 sections 6U and 4C were both pavement sections that exhibited cracking in the field. However, section 4C displayed a greater amount of cracking. All of the cracked sections also had N f values of less than 6000. This value was chosen as the critical N f value that separates the cracked sections from the uncracked. Using the value of 6000 as the critical N f value, relationships were developed between DCSE and D 1max for different m-values. These are shown in Figure 6.2 for constant values of stress and tensile strength. D 1max is defined as the D 1 value that produces an N f value of 6000 for a given DCSE f and m-value. A minimum DCSE f value of 0.75 KJ/m 3 was used for the analysis because all sections with a lower value performed poorly. 0.00E+005.00E-071.00E-061.50E-062.00E-062.50E-063.00E-063.50E-064.00E-0600.511.522.5DCSE (KJ/m3)D1max (1/psi) 0.35 0.4 0.45 0.5 0.55 0.6 m-valu e Figure 6.1. D 1max for values of DCSE and m-value

PAGE 72

02000400060008000100001200014000I75-1UI75-1CI75-2UI75-3CSR80-2USR80-1CSR 16-6USR 16-4CSR 375-1USR 375-2CTPK 1UTPK 2CNW 39-2UNW 39-1CSectionNf to Propagate 50mm 58 Figure 6.1. N f to propagate 50mm

PAGE 73

59 It was noted that the relationship presented in Figure 6.2 could be expressed using a single function of the following form: (6.1) bmDCSEaD(max)1 where D 1(max) (1/psi) is the maximum acceptable D 1 value that will achieve in good cracking performance, DCSE is the dissipated creep strain energy limit (KJ/m 3 ), m is the m-value, and a and b are regression constants. The coefficient b was determined to be 2.98, while a was determined to be a function of tensile strength and tensile stress as follows: (6.2) 810.321046.236.61099.2tSa where S t (MPa) is the ultimate tensile strength and (psi) is the tensile stress in the asphalt layer. The HMA mixture fracture toughness (K HMA ) was defined as the inverse of D 1 Therefore, a minimum fracture toughness can be defined as the inverse of D 1(max) which can be expressed as follows: DCSESmKtHMA810.3298.2(min)1046.236.61099.2 (6.3)

PAGE 74

60 0.0E+002.0E-084.0E-086.0E-088.0E-081.0E-071.2E-071.4E-0701234567St (MPa)a 100 120 150 (ps i ) Figure 6.3. Relationship between a, S t, and Furthermore, a K HMA Ratio is defined as the ratio of the mixture fracture toughness to the minimum mixture fracture toughness as follows: (min)HMAHMAHMAKKRatioK (6.4) where, K HMA equals 1/D1 and K HMA(min) is determined by Equation 6.3. The K HMA Ratio allows the comparison of the cracking resistance of different pavement sections. A mixture with a K HMA Ratio value greater than 1 will have good cracking performance, while a mixture with a K HMA ratio value of less than 1 will have poor cracking performance. Figure 6.4 shows a comparison of the field sections and their respective K HMA ratios. From the figure, it can be seen that all cracked sections exhibited a fracture toughness ratio of less than one except for those sections with a DCSE f value of less than 0.75KJ/m 3 Each of the uncracked sections exhibited a K HMA Ratio of greater than one excluding section I-10 MW1which had an unusually large DCSE f value.

PAGE 75

61 Figure 6.5 displays the K HMA ratios for all sections with DCSE f values between 0.75 KJ/m 3 and 2.5KJ/m 3 For this range of DCSE f values, it appears that the K HMA ratio is accurate in predicting the cracking performance of the pavement sections.

PAGE 76

0.000.501.001.502.002.503.00US301-BSSR80-1CI10-DEI10-DWSR 16-4CI10-MW2SR 16-6CNW 39-1CTPK 2CI75-3CI75-1CSR 375-2CUS301-BNI95-SJNI10-MW1I75-1UI75-2UTPK 1USR80-2UNW 39-2USR 375-1UI95-DNSectionKHMA Ratio Uncracked Cracke d **+note: DCSE < 0.75 KJ/ m 3 + DCSE > 2.5 KJ/ m 3 62 Figure 6.3. K HMA Ratio for all sections

PAGE 77

0.000.501.001.502.002.503.00US301-BSSR80-1CI10-DEI10-DWSR 16-4CI10-MW2SR 16-6CNW39-1CTPK2CI75-3CI75-1CSR375-2CI75-1UI75-2UTPK1USR80-2UNW39-2USR375-1UI95-DNSectionKHMA Ratio Uncracke d Cracked 63 Figure 6.4. K HMA Ratio for sections 0.75 KJ/m 3 < DCSE f < 2.5 KJ/m 3

PAGE 78

64 6.3 Mixture Properties An analysis was performed on all of the pavement sections to identify the key mixture properties that affect cracking performance. Appendix F displays the various mixture properties for each section including binder viscosity, effective asphalt content, theoretical film thickness, percent air voids, and VMA. From these figures, it is clear that no clear relationship exists between any of these properties and the relative cracking performance. However, several trends were observed when analyzing the gradations of each section. The K HMA ratios were calculated for each section with a constant stress of 120 psi. This was done to eliminate the effect of structure in the calculation. The gradations were then grouped according to high and low K HMA ratios as well as mix designation to observe any trends between gradation characteristics and K HMA ratio. These gradations are shown for all sections in Figures 6.5 through 6.8 below. Two primary differences in gradation are apparent between the low K HMA ratio sections and the high K HMA sections. The low K HMA sections move away and remain further away from the max density line. These sections either remain fine relative to the line or approach it and then gap dramatically at the finer sieves. The gradation curves for the high K HMA sections are parallel and remain close to the max density line. They also appear coarser with respect to the line without gapping drastically at the finer sieves. A summary of these trends is illustrated in Figures 6.9-6.12.

PAGE 79

0102030405060708090100Sieve sizes% Passing SR16-4C SR16-6C SR80-1C I-10 DW 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine 0.250.420.380.30KHMA Ratio 65 Figure 6.5. Gradations of Low K HMA Ratio Sections with 12.5mm Nominal Aggregate Size

PAGE 80

0102030405060708090100Sieve sizes% Passing SR 375-1U SR80-2U I-95 DN SR 375 2C NW 39 1C I 75 3C 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine 2.103.120.860.881.210.90 KHMA Ratio 66 Figure 6.6. Gradations of High K HMA Ratio Sections with 12.5mm Nominal Aggregate Size

PAGE 81

0102030405060708090100Sieve sizes% Passing US301 BS US301 BN I-10 DE I-10 MW2 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine 0.050.690.220.48KHMA Ratio 67 Figure 6.7. Gradations of Low K HMA Ratio Sections with 9.5mm Nominal Aggregate Size

PAGE 82

0102030405060708090100Sieve sizes% Passing TPK 1U NW 39-2U I75-2U I75-1U I-10 MW1 TPK 2C I75 1C 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine 1.403.890.971.781.260.910.97KHMA Ratio 68 Figure 6.8. Gradations of High K HMA Ratio Sections with 9.5mm Nominal Aggregate Size

PAGE 83

69 0102030405060708090100Sieve sizes% Passing Low KHMA High KHMA 0 200 100 50 30 16 8 4 3/8 1/2 3/4 Max DensityLine Figure 6.9. Case 1 Gradation Comparison for 12.5mm Nominal mixes 0102030405060708090100Sieve sizes% Passing Low KHMA High KHMA 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine Figure 6.10. Case 2 Gradation Comparison for 12.5mm Nominal mixes

PAGE 84

70 0102030405060708090100Sieve sizes% Passing Low KHMA High KHMA 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine Figure 6.11. Case 1 Gradation Comparison for 9.5mm Nominal mixes 0102030405060708090100Sieve sizes% Passing High KHMA Low KHMA 0 200 100 50 30 16 8 4 3/8 1/2 3/4 1 Max DensityLine Figure 6.12. Case 2 Gradation Comparison for 9.5mm Nominal mixes

PAGE 85

71 6.4 Traffic Traffic has an important effect on the initiation and propagation of surface cracks in pavements. A higher level of traffic generally means a higher number of load repetitions and a higher likelihood of high critical loads. It may be important to require higher minimum K HMA ratios for higher traffic sections during the design process. It was assumed that a road with a higher volume of traffic would require a pavement system that would provide a higher number of cycles until a crack develops. Average values of S t DCSE f m-value, and D 1 were chosen that would result in 6000 N f at a crack length of 2 in. or a K HMA Ratio of 1. The D 1 value was varied while all other values were held constant to produce different K HMA ratios. The resulting N f values were calculated using the HMA fracture mechanics model. Figure 6.13 below shows the relationship between K HMA Ratio and N f N f is directly proportional to traffic loading. Figure 6.13 shows that an increase in traffic by a factor of two will require a pavement system with two times the K HMA ratio. 00.511.522.502000400060008000100001200014000NfKHMA Ratio Figure 6.13. Relationship Between N f and Required K HMA Ratio

PAGE 86

72 From the relationship between N f and K HMA ratio, the factor F T was defined. This factor accounts for the increased K HMA ratio that would be required for increased traffic. This relationship between traffic level and required K HMA ratio was compared to actual field section data. The uncracked section with the lowest K HMA ratio in this study was section I75-1U with a value of 1.27. This section also possessed a traffic loading of approximately 500,000 ESALS/year. The relationship shown in Figure 6.13 above was calibrated to this value where at a traffic level of 500,000 ESALS/year, the F T factor equals 1.3. Figure 6.14 below shows F T vs. traffic level. 00.511.522.53020040060080010001200Traffic (ESALS/year x1000)FT Figure 6.14. FT vs. Traffic loading (ESALS/year x 1000) This calculation does not however consider the structural differences that typically exist between pavement sections with considerably different traffic loading. An analysis was performed on theoretical pavement sections with differing traffic loadings. Several pavement sections corresponding to different traffic loads were generated using the AASHTO pavement design method. These were generated for sections with traffic ranging from 50,000 ESALS/year to 1x10 6 ESALS/year. The design life of the pavement was assumed to be 20 years. Constant values of layer moduli, reliability, and standard

PAGE 87

73 deviation were assumed. A sample calculation appears in Appendix G. The layer thicknesses were calculated as well as the maximum tensile stress at the bottom of the AC layer. Using these stresses, the resulting K HMA ratios were calculated. Pavement sections with higher traffic loads resulted in thicker AC layers as well as lower stresses at the bottom of the AC layer and therefore possessed higher K HMA ratios. The factor F S considers the effect of the increase in structural capacity on the increase in K HMA ratio. The resulting relationship between traffic loading and F S is shown in Figure 6.15 below. 00.20.40.60.811.2020040060080010001200Traffic (ESALS/year x1000)FS Figure 6.15. Relationship Between Traffic Level and F S The F S factor was also calibrated to the field section I75-1U with a traffic level of approximately 500,000 ESALS/year. Figure 6.16 below shows the resulting relationship.

PAGE 88

74 0123456020040060080010001200Traffic (ESALS/year x 1000)FS Figure 6.16. F S vs. Traffic (ESALS/year x 1000) Considering the effect of traffic as well as the change in structure results in the following equation: STHMAFFquiredRatioK Re (6.5) where K HMA Ratio Required is the minimum required K HMA ratio, F T is the traffic factor, F S is the structural factor, and K HMA Ratio is the ratio of actual K HMA to the minimum required K HMA Figure 6.17 shows the minimum required K HMA ratio for different traffic loads. The value of 1 was used for sections with traffic levels less than or equal to 250k ESALS/yr.

PAGE 89

75 00.511.522.5020040060080010001200Traffic (ESALS/year x1000)Min KHMA Ratio Required Figure 6.17 Minimum K HMA Ratio Required vs. Traffic Levels

PAGE 90

CHAPTER 7 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS 7.1 Summary and Conclusions The findings of this study may be summarized as follows: The important effects of mixture properties and pavement structure on top-down cracking performance was verified. It appears that HMA fracture mechanics properly accounts for effects of mixture properties. The relative cracking performance predicted by thee HMA fracture model developed at the University of Florida agrees with field observations In some cases excessively low DCSE f appeared to control cracking. It appears that cracking develops within relatively few cycles in these cases, such that cumulative DCSE f may not appropriately represent be the mode of failure. A single parameter (K HMA (min) ) was identified and defined that allows designers to determine whether a mixture will experience top-down cracking. This parameter accounts for the effects of both mixture properties and pavement structure. It is based on M R Creep, and Strength test results. The following relationship was developed for K HMA(min) based on calibration to actual field cracking performance of mixtures: DCSESmKtHMA810.3298.2(min)1046.236.61099.2 Furthermore, a mixture fracture toughness ratio was defined that allows relative comparisons between different mixtures and pavement sections: (min)HMAHMAHMAKKRatioK where, 11DKHMA and If K Ratio > 1 No Cracking If K Ratio < 1 Cracking The K HMA Ratio was shown to predict field cracking performance for all field sections evaluated (22) except for three sections with very low or very high DCSE f 76

PAGE 91

77 Specific gradation characteristics were associated with poor cracking performance. It was concluded that the K HMA Ratio may be increased by changes in the gradation such as the degree to which the gradation curve is parallel to the max density line as well as the severity of gap grading in the finer sieves. A procedure was established to rationally account for the effects of traffic on the minimum required K HMA Ratio. The following recommendations resulted from the work: Traffic Minimum ESALS/year x 1000 K HMA Ratio <250 1 400 1.2 500 1.3 1000 1.95 7.2 Recommendations The specific condition and mechanism associated with top-down cracking is yet to be determined. The relations developed were based on relative tensile stress as determined at the bottom of the asphalt concrete layer. The critical tensile stresses at the top of the pavement may involve the introduction of a crack and a temperature gradient into the calculation as well as the residual stresses that may be induced by creep.

PAGE 92

APPENDIX A SUMMARY OF NON-DESTRUCTIVE TESTING (FWD)

PAGE 93

Table A.1: Deflection (mils0 From I-75 1U Sensor Spacing (in) Load Milepost 0 8 12 18 24 36 60 Lbs. 2.01 5.53 4.42 3.72 2.91 2.28 1.44 0.8 9142 2.02 5.95 4.89 4.19 3.22 2.49 1.5 0.76 9002 2.03 5.8 4.71 4 3.09 2.38 1.45 0.76 9086 2.04 5.98 4.82 4.11 3.17 2.43 1.54 0.85 9094 2.05 5.33 4.38 3.79 2.93 2.32 1.59 0.87 9121 2.06 5.46 4.35 3.69 2.89 2.3 1.46 0.85 8954 2.07 5.78 4.66 3.91 3.15 2.43 1.51 0.87 9089 2.08 6.09 4.91 4.2 3.25 2.5 1.6 0.89 9050 2.09 6.04 4.89 4.19 3.26 2.448 1.53 0.85 9007 2.1 5.63 4.58 3.89 3 2.36 1.52 0.86 8938 Table A.2: Deflection (mils) From I-75 1C Sensor Spacing (in) Load Milepost 0 8 12 18 24 36 60 Lbs. 1.01 6.41 5.15 4.32 3.33 2.57 1.63 1.01 8943 1.02 6.56 5.19 4.34 3.26 2.55 1.6 0.91 8803 1.03 6.28 5.01 4.21 3.15 2.46 1.62 0.94 8911 1.04 6.27 5.07 4.29 3.26 2.56 1.66 1 8891 1.05 6.38 5.17 4.37 3.36 2.65 1.69 0.96 8856 1.06 6.39 5.22 4.47 3.51 2.74 1.77 1 8771 1.07 6.3 5.16 4.4 3.47 2.77 1.79 0.95 8681 1.08 6.38 5.23 4.43 3.5 2.79 1.81 1.05 8811 1.09 6.15 5.05 4.32 3.43 2.71 1.82 1.01 8800 1.1 6.87 5.64 4.79 3.76 2.94 1.85 1 8800 79

PAGE 94

02468101214161820010203040506070Sensor Spacing (in)Deflections (mils) 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.1 80 Figure A.1 Deflections for I-75 1U

PAGE 95

02468101214161820010203040506070Sensor Spacing (in)Deflection (mils) 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.1 81 Figure A.2 Deflections for I-75 1C

PAGE 96

82 Table A.3: Deflections for I-75 2U Sensor Spacing (in) Load Milepost 0 8 12 18 24 36 60 lbs. 1.01 4.69 3.74 3.24 2.68 2.31 1.65 0.99 8883 1.02 4.39 3.5 3.03 2.5 2.08 1.5 0.91 9046 1.03 4.67 3.52 3.02 2.5 2.08 1.49 0.91 9094 1.04 4.4 3.45 2.95 2.41 2.07 1.57 0.92 8792 1.05 4.49 3.37 2.88 2.41 1.99 1.49 0.9 9022 1.06 4.73 3.81 3.28 2.72 2.25 1.66 0.91 9007 1.07 4.83 3.87 3.26 2.72 2.26 1.62 0.94 9026 1.08 4.78 3.79 3.17 2.61 2.14 1.47 0.81 8954 1.09 4.68 3.83 3.31 2.67 2.26 1.74 1.02 9054 1.1 4.64 3.68 3.09 2.57 2.18 1.61 0.93 8962 Table A.4: Deflections for I-75 3C Sensor Spacing (in) Load Milepost 0 8 12 18 24 36 60 lbs. 2.01 6.16 5.01 4.13 3.24 2.5 1.52 0.79 8943 2.02 6.16 4.85 4.11 3.2 2.52 1.62 0.8 8962 2.03 6.34 4.91 4.14 3.23 2.54 1.63 0.83 8943 2.04 5.89 4.83 4.02 3.21 2.43 1.57 0.8 8819 2.05 6.5 5.08 4.15 3.22 2.44 1.52 0.8 8927 2.06 6.69 5.41 4.58 3.59 2.8 1.75 0.82 8840 2.07 6.59 5.41 4.52 3.47 2.68 1.66 0.82 8859 2.08 6.36 5.04 4.24 3.31 2.56 1.63 0.8 8856 2.09 6.24 4.99 4.12 3.25 2.51 1.57 0.82 8803 2.1 5.94 4.72 3.93 3.07 2.44 1.56 0.82 8851

PAGE 97

02468101214161820010203040506070Sensor Spacing (in)Deflection (mils) 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.1 83 Figure A.3 Deflections for I-75 2U

PAGE 98

02468101214161820010203040506070Sensor Spacing (in)Deflection (mils) 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.1 84 Figure A.4 Deflections for I-75 3C

PAGE 99

85 Table A.5: Deflections for SR-80 2U Sensor Spacing (in) Load Milepost 0 8 12 18 24 36 60 lbs. 2.01 7.48 5.77 4.97 4.1 3.39 2.18 1 9054 2.02 7.15 5.98 5.22 4.27 3.5 2.27 1.07 8906 2.03 7.59 6.48 5.55 4.46 3.66 2.42 1.14 8800 2.04 8.55 6.7 5.53 4.41 3.59 2.35 1.13 8967 2.05 9.79 7 5.94 4.6 3.62 2.19 1.07 8864 2.06 7.25 6.18 5.36 4.37 3.52 2.24 1.14 9018 2.07 7.68 6.2 5.35 4.32 3.45 2.19 1.07 8994 2.08 8.16 6.73 5.81 4.58 3.73 2.39 1.21 8970 2.09 8.51 6.26 5.33 4.07 3.23 2.02 1.09 8720 2.1 9.35 6.71 5.64 4.38 3.5 2.16 1.11 8668 Table A.6: Deflections for SR-80 1C Sensor Spacing (in) Load Milepost 0 8 12 18 24 36 60 lbs. 1.01 10.99 7.75 5.44 3.33 2.26 1.21 0.76 8613 1.02 10.91 7.49 5.44 3.41 2.24 1.32 0.81 8450 1.03 10.31 7.02 5.07 3.02 2.07 1.19 0.78 8414 1.04 9.88 6.78 4.85 3.11 2.09 1.31 0.83 8593 1.05 11.75 8.24 5.78 3.52 2.27 1.26 0.83 8477 1.06 14.06 10.22 7.61 5.07 3.51 2.17 1.29 8315 1.07 17.67 12.91 9.74 6.54 4.56 2.72 1.51 8307 1.08 18.84 13.7 10.13 6.56 4.63 2.85 1.59 8347 1.09 17.02 12.54 9.62 6.68 4.85 2.91 1.63 8334 1.1 17.12 13.12 10.34 7.44 5.37 3.22 1.71 8318

PAGE 100

02468101214161820010203040506070Sensor Spacing (in)Deflection (mils) 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.1 86 Figure A.5 Deflections for SR-80 2U

PAGE 101

02468101214161820010203040506070Sensor Spacing (in)Deflection (mils) 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.1 87 Figure A.6 Deflections for SR-80 1C

PAGE 102

88 012345670102030405060Sensor Spacing (in)Deflection (mils) 70 Measured Computed 0123456701020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.8 Measured and Computed Deflections for I-75 1U Location 06 012345670102030405060Sensor Spacing (in)Deflection (mils) 70 Measured Computed Figure A.9 Measured and Computed Deflections for I-75 1U Location 09 Figure A.7 Measured and Computed Deflections for I-75 1U Location 03

PAGE 103

89 01234567801020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.10 Measured and Computed Deflections for I-75 1C Location 03 01234567801020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.11 Measured and Computed Deflections for I-75 1C Location 08 0123456780102030405060Sensor Spacing (in)Deflection (mils) 70 Measured Computed Figure A.12 Measured and Computed Deflections for I-75 1C Location 10

PAGE 104

90 01234567801020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.13 Measured and Computed Deflections for I-75 2U Location 02 0123456780102030405060Sensor Spacing (in)Deflection (mils) 70 Measured Computed Figure A.14 Measured and Computed Deflections for I-75 2U Location 09 0123456780102030405060Sensor Spacing (in)Deflection (mils) 70 Measured Computed Figure A.15 Measured and Computed Deflections for I-75 2U Location 02

PAGE 105

91 0123456780102030405060Sensor Spacing (in)Deflection (mils) 70 Measured Computed 01234567801020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.17 Measured and Computed Deflections for I-75 1C Location 04 01234567801020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.18 Measured and Computed Deflections for I-75 1C Location 06 Figure A.16 Measured and Computed Deflections for I-75 1C Location 03

PAGE 106

92 02468101201020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.19 Measured and Computed Deflections for SR-80 2U Location 05 02468101201020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.20 Measured and Computed Deflections for SR-80 2U Location 09 02468101201020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.21 Measured and Computed Deflections for SR-80 2U Location 01

PAGE 107

93 02468101201020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed 02468101201020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.23 Measured and Computed Deflections for SR-80 1C Location 03 02468101201020304050607Sensor Spacing (in)Deflection (mils) 0 Measured Computed Figure A.23 Measured and Computed Deflections for SR-80 1C Location 03 Figure A.22 Measured and Computed Deflections for SR-80 1C Location 02

PAGE 108

APPENDIX B SUMMARY OF FDOT FLEXIBLE PAVEMENT CONDITION SURVEY

PAGE 109

0246810121990199219941996199820002002YearCrack Rating I75-1U I75-1C Figure B.1 Cracking Ratings from I-75 1U and 1C 0246810121990199219941996199820002002YearCrack Rating I75-2U I75-3C Figure B.2 Cracking Ratings from I-75 2U and 3C 95

PAGE 110

96 0246810121990199219941996199820002002YearCrack Rating SR 80-2U SR 80-1C Figure B.3 Cracking Ratings from SR-80 2U and 1C

PAGE 111

APPENDIX C SUMMARY OF VOLUMETRIC PROPERTIES

PAGE 112

98 Table C.1 Bulk Specific Gravity for each Sample I75-1U I75-1C I75-2U I75-3C SR 80-2U SR 80-1C Sample WP BWP WP BWP WP BWP WP BWP WP BWP WP BWP 1 2.329 2.294 2.310 2.268 2.327 2.254 2.296 2.235 2.274 2.211 2.278 2.222 2 2.326 2.281 2.310 2.258 2.297 2.241 2.293 2.233 2.258 2.206 2.277 2.222 3 2.326 2.280 2.308 2.253 2.295 2.240 2.291 2.228 2.253 2.202 2.276 2.219 4 2.326 2.290 2.280 2.308 2.252 2.289 2.235 2.225 2.252 2.197 2.275 2.214 5 2.325 2.279 2.306 2.248 2.286 2.233 2.289 2.218 2.252 2.195 2.274 2.213 6 2.325 2.277 2.305 2.247 2.286 2.232 2.287 2.218 2.250 2.188 2.274 2.213 7 2.325 2.277 2.304 2.246 2.284 2.231 2.286 2.216 2.246 2.188 2.273 2.210 8 2.324 2.275 2.303 2.242 2.282 2.226 2.283 2.213 2.238 2.179 2.272 2.207 9 2.324 2.274 2.303 2.242 2.282 2.226 2.281 2.212 2.237 2.178 2.271 2.206 10 2.324 2.274 2.302 2.238 2.278 2.218 2.277 2.212 2.236 2.177 2.269 2.202 11 2.323 2.274 2.302 2.237 2.278 2.217 2.277 2.210 2.227 2.175 2.269 2.201 12 2.323 2.270 2.301 2.236 2.277 2.216 2.277 2.206 2.226 2.175 2.269 2.200 13 2.322 2.269 2.300 2.234 2.276 2.214 2.272 2.205 2.224 2.174 2.265 2.200 14 2.322 2.267 2.300 2.232 2.275 2.214 2.262 2.202 2.222 2.174 2.262 2.199 15 2.319 2.266 2.300 2.232 2.273 2.209 2.252 2.190 2.206 2.171 2.261 2.196 16 2.318 2.264 2.296 2.231 2.272 2.200 2.190 2.200 2.167 2.259 2.194 17 2.317 2.262 2.295 2.230 2.267 2.194 2.173 2.199 2.160 2.256 2.181 18 2.259 2.215 2.170 2.152 2.170 2.148 2.222 2.176 Average 2.323 2.274 2.303 2.241 2.284 2.221 2.281 2.208 2.232 2.181 2.267 2.204 St. Dev 0.003 0.008 0.004 0.012 0.014 0.020 0.012 0.021 0.026 0.016 0.013 0.013

PAGE 113

99 Table C.2 Effective Asphalt Content, Film Thickness and VMA Factors Gradation Size m2/kg ft2/lb I75 1C I75 1U I75 2 I75 3 SR80 1 SR80 2 19 0.41 2 100.00 10 0.00 10 0.00 100.00 100.00 100.00 12.5 0.41 2 99.46 97 .69 97 .74 90.75 92.68 93.63 9.5 0.41 2 95.74 93 .71 92 .47 79.54 83.57 84.79 4.75 0.41 2 75.92 74 .58 68 .90 58.33 60.48 64.41 2.36 0.82 4 57.68 55 .58 50 .47 47.54 45.51 52.38 1.18 1.64 8 46.75 44 .24 41 .87 42.09 35.66 41.05 0.06 2.87 14 38.67 36 .78 35 .18 37.11 28.47 32.69 0.03 6.14 30 27.36 25 .61 22 .58 25.77 21.76 25.65 0.015 12.29 60 12.83 12 .30 10 .67 12.99 11.51 14.60 0.0075 32.77 160 5.86 5. 58 5. 02 5.88 6.19 6.18 Surface area Size I75-1C I75 -1U I75 -2U I75-3C SR80 1 SR80 2 19 0.410 0. 410 0. 410 0.410 0.410 0.410 12.5 0.408 0. 401 0. 401 0.372 0.380 0.384 9.5 0.393 0. 384 0. 379 0.326 0.343 0.348 4.75 0.311 0. 306 0. 282 0.239 0.248 0.264 2.36 0.473 0. 456 0. 414 0.390 0.373 0.430 1.18 0.767 0. 726 0. 687 0.690 0.585 0.673 0.06 1.110 1. 056 1. 010 1.065 0.817 0.938 0.03 1.680 1. 573 1. 386 1.582 1.336 1.575 0.015 1.576 1. 512 1. 311 1.596 1.415 1.794 0.0075 1.920 1. 829 1. 645 1.927 2.028 2.025 G se 2.61 2. 57 2. 61 2.62 2.56 2.60 Surface area, m 2 /kg 9.047 8. 651 7. 925 8.598 7.935 8.841 Total AC 6.78 6. 20 6. 15 6.640 6.450 6.760 G sb 2.490 2. 476 2. 462 2.436 2.407 2.433 Effective AC, % 4.92 4. 83 3. 90 4.41 3.84 4.32 Gmm 2.369 2. 349 2. 386 2.380 2.338 2.359 G mb 2.242 2. 274 2. 241 2.190 2.213 2.197 P b 1.90 1. 45 2. 37 3.01 2.57 2.74 Weight of Absorbed AC, g 39.8 30 .8 49.88 61.57 53.18 56.03 Effective Vol. Of AC, ml 107.83 10 6.07 83.90 79.62 85.37 88.16 Vol. Of AC, ml 147.61 13 6.91 13 3.78 141.19 138.55 144.19 Film thickness, micro-meters 5.70 5.75 5.03 4.53 5.20 4.87 VMA 16.05 13 .84 14 .58 16.06 14.00 15.80

PAGE 114

APPENDIX D MIXTURE TEST RESULTS

PAGE 115

Table D.1 Resilient Modulus Total Resilient Modulus (GPa) Cycle Section Temp 1 2 3 Average 0 14.76 14.81 15.24 14.94 I75-1U 10 10.92 10.88 10.93 10.91 20 7.68 7.61 7.65 7.65 0 16.83 16.63 17.18 16.88 I75-1C 10 11.28 10.96 11.17 11.14 20 8.08 7.62 7.56 7.75 0 18.42 18.56 18.23 18.40 I75-2U 10 10.18 10.42 10.28 10.29 20 7.77 7.64 7.64 7.68 0 16.79 16.76 17.04 16.86 I75-3C 10 11.52 11.58 11.64 11.58 20 8.19 8 7.98 8.06 0 17.78 18.4 18.06 18.08 SR80-2U 10 13.6 13.5 13.24 13.45 20 11.1 10.86 11.06 11.01 0 18.24 19.09 18.14 18.49 SR80-1C 10 13.41 13.42 13.33 13.39 20 9.22 9.21 8.96 9.13 Table D.2 Tensile Strength Tensile Strength (MPa) Poisson's Specimen Section Temp Ratio 1 2 3 Average 0 0.34 3.00 3.20 2.69 2.96 I75-1U 10 0.38 2.29 1.83 1.92 2.01 20 0.41 1.27 1.25 1.24 1.25 0 0.38 2.77 2.77 2.34 2.63 I75-1C 10 0.46 1.61 1.60 1.73 1.65 20 0.46 1.04 0.88 1.19 1.04 0 0.37 3.38 3.42 2.90 3.23 I75-2U 10 0.34 2.03 1.88 1.77 1.89 20 0.44 1.28 1.23 1.30 1.27 0 0.38 2.92 2.54 2.33 2.60 I75-3C 10 0.49 1.56 1.87 1.60 1.68 20 0.38 1.34 1.40 1.37 1.37 0 0.23 2.71 2.96 2.82 2.83 SR80-2U 10 0.34 2.35 2.22 2.61 2.39 20 0.46 1.83 1.66 1.17 1.55 0 0.39 2.33 2.45 2.27 2.35 SR80-1C 10 0.39 1.60 1.58 1.59 1.59 20 0.32 1.21 1.02 1.82 1.35 101

PAGE 116

102 Table D.3 Creep Compliance Creep Compliance (1/Gpa) Poisson's 1 10 100 Section Temp Ratio (sec) (sec) (sec) I75-1U 0 0.46 0.095 0.155 0.324 10 0.39 0.196 0.420 1.029 20 0.45 0.389 1.026 3.002 I75-1C 0 0.45 0.076 0.128 0.268 10 0.47 0.202 0.420 1.060 20 0.43 0.358 0.881 2.466 I75-2U 0 0.45 0.070 0.117 0.239 10 0.41 0.216 0.436 1.143 20 0.43 0.431 1.153 3.253 I75-3C 0 0.42 0.084 0.132 0.266 10 0.45 0.152 0.296 0.766 20 0.32 0.351 1.100 3.252 SR80-2U 0 0.30 0.071 0.091 0.126 10 0.39 0.103 0.164 0.324 20 0.44 0.194 0.349 0.794 SR80-1C 0 0.48 0.060 0.087 0.144 10 0.45 0.104 0.214 0.540 20 0.35 0.269 0.572 1.523

PAGE 117

103 Table D.4 M-value, Failure Stra in, Fracture Energy, DCSE, Initia l Tangent Modulus, D0, and D1 Failure Fracture Strain Energy DCSE Initial Tangent Section Temp M-value (Microstrain) (KJ/m 3 ) (KJ/m 3 ) Modulus D0 D1 I75-1U 0 0.4366 794.77 1.5 1.207 7 4.13E-07 2.43E-07 10 0.4363 1437.44 2.0 1.814 3.7 4.72E-07 8.88E-07 20 0.5435 2810.96 2.5 2.398 1.5 1.36E-06 1.59E-06 I75-1C 0 0.4314 582.84 0.8 0.595 7.1 3.21E-07 2.07E-07 10 0.4713 1028.05 1.1 0.978 3.8 6.55E-07 7.55E-07 20 0.4648 1363.28 1.0 0.930 2.1 6.87E-07 1.89E-06 I75-2U 0 0.4187 822.53 1.7 1.416 7.6 2.87E-07 2.87E-07 10 0.5173 1066.72 1.3 1.126 4.2 8.54E-07 6.50E-07 20 0.4689 2663.56 2.5 2.395 1.7 4.75E-07 2.53E-06 I75-3C 0 0.4479 617.21 0.9 0.700 6.5 4.00E-07 1.82E-07 10 0.5192 715.74 0.8 0.678 5 6.48E-07 4.21E-07 20 0.4697 2715.98 2.7 2.584 1.9 -1.71E-14 2.57E-06 SR80-2U 0 0.3088 444.92 0.7 0.479 9 3.70E-07 1.22E-07 10 0.4176 679.15 1.0 0.787 6.3 4.66E-07 2.54E-07 20 0.4584 962.22 1.0 0.891 3.7 7.81E-07 5.65E-07 SR80-1C 0 0.3216 424.49 0.5 0.351 6.8 2.52E-07 1.66E-07 10 0.4797 495.27 0.3 0.206 3.6 3.63E-07 3.66E-07 20 0.5131 849.77 0.7 0.600 1.7 9.95E-07 8.93E-07

PAGE 118

104 Table D.5 Dissipated Creep Strain Energy Calculation Section Temp MR e f S t e O EE FE DCSE I75-1U 0 14.94 794.77 2.96 794.57 0.29 1.5 1.207 10 10.91 1437.44 2.01 1437.26 0.19 2.0 1.814 20 7.65 2810.9 6 1.25 2810.80 0.10 2.5 2.398 I75-1C 0 16.88 582.84 2.63 582.68 0.20 0.8 0.595 10 11.14 1028.05 1.65 1027.90 0.12 1.1 0.978 20 7.75 1363.2 8 1.04 1363.15 0.07 1.0 0.930 I75-2U 0 18.4 822.53 3.23 822.35 0.28 1.7 1.416 10 10.29 1066.72 1.89 1066.54 0.17 1.3 1.126 20 7.68 2663.5 6 1.27 2663.39 0.10 2.500 2.395 I75-3C 0 16.86 617.21 2.60 617.06 0.20 0.9 0.700 10 11.58 715.74 1.68 715.59 0.12 0.8 0.678 20 8.06 2715.9 8 1.37 2715.81 0.12 2.7 2.584 SR80-2U 0 18.08 444.92 2.83 444.76 0.22 0.7 0.479 10 13.45 679.15 2.39 678.97 0.21 1.0 0.787 20 11.01 962.22 1.55 962.08 0.11 1.0 0.891 SR80-1C 0 18.49 424.49 2.35 424.36 0.15 0.5 0.351 10 13.39 495.27 1.59 495.15 0.09 0.3 0.206 20 9.13 849.77 1.35 849.62 0.10 0.7 0.600

PAGE 119

APPENDIX E SUMMARY OF CRACK GROWTH MODEL RESULTS

PAGE 120

Table E.1. Estimated Loading Stresses (psi) Section Temp 7000 LB 9000 LB 11000 LB 0 134 172 211 I75-1U 10 119 153 188 20 102 131 160 0 138 177 216 I75-1C 10 119 152 186 20 103 132 161 0 99.3 128 156 I75-2U 10 77.6 99.8 122 20 66.3 85.2 104 0 139 178 218 I75-3C 10 122 157 191 20 106 137 167 0 149 191 234 SR80-2U 10 136 175 214 20 125 161 197 0 287 369 451 SR80-1C 10 247 318 389 20 206 264 323 106

PAGE 121

107 Table E.2. N f to Initiation for DCSE DCSE N f to Failure (Initiation) Temp 7000 9000 11000 0 24598 14930 9921 I75-1U 10 12854 7776 5150 20 6329 3837 2572 0 13912 8457 5679 I75-1C 10 6422 3936 2629 20 3402 2072 1392 0 50381 30321 20413 I75-2U 10 14888 9001 6023 20 15366 9305 6245 0 16379 9988 6659 I75-3C 10 5531 3340 2256 20 6351 3802 2559 0 40051 24373 16239 SR80-2U 10 16998 10266 6865 20 7731 4660 3113 0 6462 3909 2617 SR80-1C 10 738 444 298 20 990 600 402 Table E.3. N f to Initiation for FE FE N f to Failure (Initiation) Temp 7000 9000 11000 0 29986 17971 11746 I75-1U 10 13953 8354 5459 20 6513 3915 2596 0 18078 10743 7008 I75-1C 10 7025 4229 2758 20 3539 2108 1378 0 60032 35949 24054 I75-2U 10 17005 10208 6770 20 15952 9625 6431 0 20422 12204 7924 I75-3C 10 6277 3691 2413 20 6554 3891 2592 0 56088 33177 21289 SR80-2U 10 20892 12338 8017 20 8384 4938 3201 0 7617 3980 2139 SR80-1C 10 752 325 112 20 1000 545 314

PAGE 122

108 Table E.4. N f to Propagate 50mm Temp 7000 9000 11000 0 19254 12694 9083 I75-1U 10 11055 7272 4990 20 5990 3779 2568 0 11438 7564 5380 I75-1C 10 5933 3815 2605 20 3325 2068 1587 0 34550 22659 16382 I75-2U 10 11290 7457 5321 20 12656 8382 5900 0 13548 9076 6333 I75-3C 10 5090 3235 2238 20 5954 3721 2545 0 33063 22078 15438 SR80-2U 10 14368 9506 6629 20 7312 4581 3104 0 4155 3733 3733 SR80-1C 10 703 703 703 20 933 933 933

PAGE 123

APPENDIX F SUMMARY OF SECTION DATA FOR ALL SECTIONS

PAGE 124

Table F.1. Summary of Traffic Loading (Annual ESALS in 1000) Section ESALS/YR Section ESALS/YR SR 16-4C 21 I75-1C 573 SR 16-6U 21 I75-2U 576 SR 375-2C 68 I75-3C 674 SR 375-1U 76 I-10 MW1 546 TPK 2C 166 I-95 DN 1192 NW 39-2U 182 I-95SJN 1192 NW 39-1C 190 I-10 DE 681 TPK 1U 195 I-10 DW 681 SR80-2U 207 I-10 MW2 546 SR80-1C 221 US 301BN 558 I75-1U 558 US 301BS 558 Table F.2. Summary of Loading Stress (psi) Section Stress Section Stress I75-1U 153 TPK 1U 116 I75-1C 152 TPK 2C 162 I75-2U 99.8 NW 39-2U 208 I75-3C 157 NW 39-1C 282 SR80-2U 175 I10-MW1 158 SR80-1C 318 I10-MW2 133 SR 16-6U 120 US301-BN 98.2 SR 16-4C 120 US301-BS 128 US 19-1U 87 I95-DN 53.9 US 19-2C 146 I95-SJN 92.1 SR 375-1U 120 I10-DW 138 SR 375-2C 120 I10-DE 120 110

PAGE 125

111 Table F.3. Summary of Mixture Test Results Needed for Cracking Model Section m-value FE DCSE St D1 (KJ/m 3 ) (KJ/m 3 ) (MPa) (1/psi) I75-1U 0.42 2.0 1.81 2.01 9.89E-07 I75-1C 0.43 1.1 0.98 1.65 9.73E-07 I75-2U 0.44 1.3 1.13 1.89 9.99E-07 I75-3C 0.45 0.8 0.68 1.68 6.23E-07 SR80-2U 0.36 1.0 0.79 2.39 3.52E-07 SR80-1C 0.47 0.3 0.21 1.59 3.82E-07 SR 16-6U 0.49 0.5 0.41 1.24 6.65E-07 SR 16-4C 0.61 0.3 0.25 0.89 3.66E-07 US 19-1U 0.6 1.7 1.55 1.53 8.29E-07 US 19-2C 0.58 3.3 1.91 2.73 4.11E-07 SR 375-1U 2.12 0.68 2.15 TPK 1U 0.96 1.3 1.14 1.7 NW 39-2U 1.75 NW 39-1C 0.23 2.4 2.26 1.82 I10-MW2 0.59 1.05 US301-BN 0.37 0.3 0.27 1.13 6.32E-07 US301-BS 0.39 0.1 0.03 0.93 8.47E-07 I95-DN 0.48 1.2 1.20 1.3 9.59E-07 I95-SJN 0.51 0.8 0.53 1.94 3.60E-07 I10-DW 0.46 0.8 0.66 1.33 1.67E-06 I10-DE 0.52 1.0 0.88 1.25 2.21E-06 0.52 2.3 2.59 4.63E-07 SR 375-2C 1.7 1.53 3.80E-07 0.47 1.1 1.96 4.67E-07 TPK 2C 0.49 8.10E-07 0.45 2.0 2.65 3.14E-07 0.54 0.3 1.41 9.18E-07 I10-MW1 0.55 8.31E-07 1.1 1.18 8.67E-07 Table F.4. Number of Cycles to Propagate Crack Length of 50mm Section Nf to Prop I75-1U 7410 I75-1C 4039 I75-2U 8311 I75-3C 3543 SR80-2U 10147 SR80-1C 704 SR 16-6U 2610 SR 16-4C 1376 US 19-1U 6530 US 19-2C 6131 SR 375-1U 12827 SR 375-2C 4314 TPK 1U 9250 TPK 2C 3328 NW 39-2U 9903 NW 39-1C 2628

PAGE 126

112 01234567I-10 MW1US 19-1USR 375-1UTPK 1UNW 39-2UI-95 DNSR 16-6UI75-1UI75-2USR-80 2UI-95SJNI-10 DEI-10 DWSR 375-2CTPK 2CI-10 MW2US 301 BNUS 301 BSSR 16-4CUS 19-2CNW 39-1CI75-1CI75-3CSR-80 1CSectionsEffective Asphalt Content % UncrackedCracked Figure F.1. Effective Asphalt Content (%)

PAGE 127

024681012I-10 MW1US 19-1USR 375-1UTPK 1UNW 39-2UI-95 DNSR 16-6UI75-1UI75-2USR 80-2U I-10 DEI-10 DWSR 375-2CTPK 2CI-10 MW2US 301 BNUS 301 BSSR 16-4CUS 19-2CNW 39-1CI-95SJNI75-1CI75-3CSR 80-1C Section% Air Voids WP BWP CrackedUncracked Figure F.2 Percent Air Voids (%) 113

PAGE 128

0.02.04.06.08.010.012.0I-10 MW1US 19-1USR 375-1UTPK 1UNW 39-2UI-95 DNSR 16-6UI75-1UI75-2USR-80 2UI-95SJNI-10 DEI-10 DWSR 375-2CTPK 2CI-10 MW2US 301 BNUS 301 BSSR 16-4CUS 19-2CNW 39-1CI75-1CI75-3CSR-80 1CSectionFilm Thickness, microns UncrackedCracked Figure F.3. Theoretical Film Thickness (microns) 114

PAGE 129

02468101214161820I-10 MW1US 19-1USR 375-1UTPK 1UNW 39-2UI-95 DNSR 16-6UI75-1U I75-2U SR 80-2U I-95SJNI-10 DEI-10 DWSR 375-2CTPK 2CI-10 MW2US 301 BNUS 301 BSSR 16-4CUS 19-2CNW 39-1CI75-1C I75-3C SR 80-1C SectionVMA (%) UncrackedCracked Figure F.4. VMA (%) 115

PAGE 130

116 LIST OF REFERENCES Ewalds, H.L., and R.J.H. Wanhill, Fracture M echanics, Delftse Uitgevers Maatschappij, Delft,Netherlands, and Edward Arnold Publishers, London, 1986. Garcia, O.F., Asphalt Mixture and Loading Effects on Surface-Cracking of Pavements, Masters Thesis, University of Florida, Gainesville, 2002. Honeycutt, K.E., Effect of Gradation a nd other Mixture Properties on the Cracking Resistance of Asphalt Mixtures, Master s Thesis, University of Florida, Gainesville, 2000. Huang, Y.H., Pavement Analysis and Design, Prentice Hall, Englewood Cliffs NJ, 1993. Jacobs, M.M.J., Crack Growth in Asphalti c Mixes, Ph.D. Dissertation, Delft, The Netherlands, Nelft University of Technology, 1995. Kandhal P.S. and S. Chakraborty, Evaluati on of Voids in the Mineral Aggregates, NCAT Report No. 96-4, Na tional Center for Asphalt Technology, March 1996. Malan, G.W., P.J. Strauss and F. Hugo, A Field Study of Premature Surface Cracking in Asphalt, Proceedings of the Association of Asphalt Paving Technologists, Vol. 58, pp. 142-162, 1989. Monismith, C.L., Fatigue Characteristics of Asphalt Paving Mixtures and Their Use in Pavement Design, Proceedings of the Eighteen Paving Conference, Vol. 54, pp,124-132, 1981. Monismith, C.L., J.A. Epps, and F.N. Finn, Improved Asphalt Mix Design, Proceedings of the Association of Asphalt Paving Technologists, Vol. 55, pp. 124132, 1985. Myers, L.A., Mechanism of Wheel Path Cr acking That Initiates at the Surface of Asphalt Pavements, Masters Thesis, Univ ersity of Florida, Gainesville, 1997. Myers, L.A., Development and Propagati on of Surface-Initiate d Longitudinal Wheel Path Cracks in Flexible Highway Pavement s, Ph.D. Dissertation, University of Florida, Gainesville, 2000. Nukunya, B. Evaluation of Aggregate Type, Gradation and Volumetric Properties for Design and Acceptance of Durable Superp ave Mixtures, Ph.D. Dissertation, University of Florida, Gainesville, 2001.

PAGE 131

117 Pell, P. and F. Taylor, Asphaltic Road Materials in Fatigue, Proceedings of the Association of Asphalt Paving Technologists, Vol. 38, pp. 371-421, 1969. Roque, R., W.G. Buttlar, B.E. Ruth, M. Tia, S.W. Dickison, and B. Reid, Evaluation of SHRP Indirect Tension Tester to Mitigate Cracking in Asphalt Pavements and Overlays, Final Report to the Florida Department of Transportation, University of Florida, Gainesville, 1997. Sedwick, S.C., Effect of Asphalt Mixture Properties and Characteristics on Surface-Initiated Longitudinal Wheel Path Cracking, Masters Thesis, University of Florida, Gainesville, 1998. Sousa, J.B., G. Way and J. Harvey, Performance Based Mix Design and Field Quality Control for Asphalt-Aggregate Overlays, Transportation Research Board, Record No. 1543, pp. 46-62, 1996. Valkering C.P. and G. Van Gooswilligen, The Role of the Binder Content in the Performance-Related Properties of Asphaltic Mixes for Surface Layers, Proceedings of the Association of Asphalt Paving Technologists, Vol. 58, pp.238-255, 1989. Zhang, Z., Identification of Suitable Crack Growth Law for Asphalt Mixtures Using the Superpave Indirect Tensile Test (IDT), Ph.D. Dissertation, University of Florida, Gainesville, 2000.

PAGE 132

BIOGRAPHICAL SKETCH Adam Jajliardo was born in Manchester, Connecticut on November 21, 1978. He graduated from High School from Woodstock Academy in 1997. Adam attended The University of Connecticut and received a Bachelor of Science in Civil Engineering degree in 2001. Adam gained admission to the University of Florida in 2001 and began his studies for his Master of Engineering degree. 118


Permanent Link: http://ufdc.ufl.edu/UFE0000706/00001

Material Information

Title: Development of specification criteria to mitigate top-down cracking
Physical Description: Mixed Material
Creator: Jajliardo, Adam Paul ( Author, Primary )
Publication Date: 2003
Copyright Date: 2003

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000706:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000706/00001

Material Information

Title: Development of specification criteria to mitigate top-down cracking
Physical Description: Mixed Material
Creator: Jajliardo, Adam Paul ( Author, Primary )
Publication Date: 2003
Copyright Date: 2003

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000706:00001


This item has the following downloads:


Full Text












DEVELOPMENT OF SPECIFICATION CRITERIA TO
MITIGATE TOP-DOWN CRACKING


















By

ADAM PAUL JAJLIARDO


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Adam Paul Jajliardo















ACKNOWLEDGMENTS

I would like to first thank my advisor and my committee chairman, Dr. Reynaldo

Roque for his advice, guidance and support. Without his technical and personal

expertise, this would not have been possible. I would also like to acknowledge my other

committee members, Dr. Bjorn Birgisson and Dr. Mang Tia, who have lent their

knowledge and experience.

Special thanks go to Mr. George Lopp for his support in the laboratory and his

valuable advice. My deepest thanks go to all the members of the Civil Engineering

materials group for their friendship and support during the past two years. They include

Tait Karlson, Oscar Garcia, Tipakorn Samarnrak, Jagannatha Katkuri, Claude Villiers,

Jeff Frank, JaeSeung Kim, SungHo Kim, Booil Kim, and Boonchi Sangpetngam.

I would like to express a very sincere appreciation to my wife Wendy for her love,

support, and friendship. I would also like to thank all my family and friends back home

who have also supported me during this time.
















TABLE OF CONTENTS
Page

A C K N O W L E D G M E N T S ......... .................................................................................... iii

LIST OF TABLES ....................................................... ............ .............. .. vii

LIST OF FIGURES ......... ......................... ...... ........ ............ ix

A B STR A C T ........... .................................. ....... ..................................xiii

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

1.1 B background ......... ...... ................................................................... ........... 1
1.2 O bjectiv es ...................... .. ............. .. ........... .................................. . 2
1.3 Scope..................................................... . 2
1.4 Research A approach ........ ...... ..................... ............ .. .... .. ...... .. ..

2 LITER A TU R E REV IEW ............................................................. ....................... 4

2.1 Fracture in A sphalt Pavem ents ........................................... .......................... 4
2.2 Mechanisms of Fracture in Asphalt Pavements ........... .....................................4
2.2.1 Traditional Fatigue A pproach.................................... ....................... 4
2.2.2 Fracture Mechanics Method ............... ................ ...............6
2.2.3 Dissipated Creep Strain Energy......................................... ............... 7
2.3 Mixture Properties Related to Fatigue Resistance ........................... ...............8
2.3.1 M ixture Stiffness ........................ ....... .. ..... .............. ..8
2.3.2 A ir V oid Content .......... .. .... ................ ......................... ... ...... 9
2.3.3 Voids in the Mineral Aggregate (VMA) ............................................... 9
2.3.4 Asphalt Content and Theoretical Film Thickness .......................................9
2.3.5 B inder V iscosity ......................................... .......... .... .. .. ........ .... 10
2.3.6 A ggregate G radiation ........................................................ ............... 11
2 .4 P rev iou s Stu dies........... ..... ..................................................................... .. .... 11
2 .5 S u m m a ry ............................................................................................12

3 DESCRIPTION OF TEST SECTIONS................................ ........................ 14

3.1 Locations and A ge ........................................................................ ... 14
3.2 Pavem ent Structure .............................................. ............. .... ......... 15
3.3 Traffic Volume ............................... ... ..... .. ...... ............... 16









3.4 Environmental Conditions ...................... ................. .... ............16
3.5 Perform ance of the Sections ......... .. ..... .. ........................... ............... 16
3 .5 .1 O v erv iew ..............................................................16
3.5.2 Field O observations .......... .. .. ... .......... .. .... ....... .......... 17

4 M ATERIALS AND M ETHOD S ........................................ ......................... 21

4.1 Extraction of the Field Cores............. .. .............. ......... ................. .... 21
4.2 Measuring and Cutting the Field Cores.................................................. 22
4.3 Selecting Sam ples for Testing ........................................ ......................... 22
4.4 Crack Rating .............. ......... ............... 23
4.5 M mixture Testing ............................. ..... ..... .... .. .......... .............. 24
4.6 Asphalt Extractions and Binder Testing ....................................................25
4.7 A aggregate Tests ...................... .................... ................... .... ....... 25
4.8 V olum etric Properties ........... .................... ........ ................ ............... 26

5 ANALYSIS AND FINDINGS ............................................................................29

5.1 Volumetric Properties and Extraction-Recovery Results...............................29
5.11 A ir V oid C ontent ........................................... ........ .......... .. ........ .... 29
5.1.2 E effective A sphalt C ontent ................................... ..................... .. .......... 31
5.1.3 A ggregate G radiation ............................................................................ 31
5.1.4 Theoretical Film Thickness ..................................................................... 36
5.1.5 B inder V iscosity ................... .............. ............... .... ........37
5.2 M mixture R results ............... ................. ........................... ..... ... ... 37
5.2 .1 R esilient M odulu s............ .................................. ........ ........ ........... 38
5 .2 .2 C reep C om plian ce ........................................................... .....................39
5 .2 .3 T en sile Strength ............ .... .......................................... ........ .... .......... 4 0
5.2 .4 F failure Strain ......................... .. .................... ......... ........... 40
5.2.5 m -value ........................................................... ..... ....................4 1
5.2.6 Fracture Energy Density and Dissipated Creep Strain Energy .................42
5.3 N on-D destructive Testing (FW D ) ........................................ ...................... 44
5.3.1 Pavem ent Structures ..................................................... ...................44
5.3.2 Loading Stresses ................................................ .... .. .. ................. 46
5.4 C rack G row th M odel ................................................................. .....................47
5.5 Individual Analysis of the Sections ....................................................................53
5.5.1 1-75 lU and 1C .......................................... ........... ... ........ 53
5.5.2 1-75 2U and 3C ........................ .... ................ ... .... .. ........... 54
5.2.3 SR-80 2U and 1C...................................... .......... .. ............. 54

6 FU R TH ER A N A L Y SIS ............................................................. ....................... 56

6.1 Section D ata and M ixture Test R results ........................................ .....................56
6.2 M ixture Fracture Toughness.......................................... ........................... 57
6 .3 M ixtu re P rop erties ........................................................................ .................. 64
6 .4 T ra ffi c .............................................................................7 1



v









7 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS ...........................76

7.1 Sum m ary and Conclusions ............................................................................76
7 .2 R ecom m endation s......................................................................... .................. 77

APPENDIX

A SUMMARY OF NON-DESTRUCTIVE TESTING (FWD) ...................................78

B SUMMARY OF FDOT FLEXIBLE PAVEMENT CONDITION SURVEY ...........94

C SUMMARY OF VOLUMETRIC PROPERTIES ................................................97

D M IX TU R E TE ST R E SU L TS ........................................................ ..................... 100

E SUMMARY OF CRACK GROWTH MODEL RESULTS .................................... 105

F SUMMARY OF SECTION DATA FOR ALL SECTIONS ................................... 109

L IST O F R E F E R E N C E S ............................ ......................................... ..................... 116

BIOGRAPH ICAL SKETCH .............. ........................ .................... ............... 118
















LIST OF TABLES

Table pge

3.1 L location of the Sections ............................................................................ .... .. 14

3.2 A ge of the Sections ........................... ................ ................... .... .. ... 15

3.3 T hickness of the layers (in) ......................................................................... ... ... 15

3.4 Layer M oduli for each Section (ksi) ............................................. ............... 16

3.5 Traffic Volumes for each Section (Millions) .........................................................16

4.1 A average thickness ............................................................... .... ...... 23

4.2 A average Bulk Specific G ravity ........................................ .......................... 23

4.3 Cracking Criteria (After Sedwick, 1998) ...................................... ............... 24

4 .4 C ra ck R atin g s ..................................................................................................... 2 4

5.1 A ir void content.................................................... 30

5.2 Layer M oduli from FW D Analysis ........................................ ....... ............... 45

5.3 E 1/E2 R atios................................................... 46

A .1 D election (m ilsO From 1-75 1U ...................................... .......................... 79

A .2 D election (m ils) From 1-75 C ........................................ .......... ............... 79

A .3 D elections for 1-75 2U .................................................. .............................. 82

A .4 D elections for 1-75 3C ................................................. ............................... 82

A .5 D elections for SR -80 2U ................................................ ............................. 85

A .6 D elections for SR -80 1C ...................................... ............................................85

C.1 Bulk Specific Gravity for each Sample......................................... ............... 98

C.2 Effective Asphalt Content, Film Thickness and VMA ........................................99









D .1 R resilient M odulus .................... ................. ........................................... 101

D.2 Tensile Strength................ ........ ..................... ........ 101

D .3 C reep C om p lian ce ........................... .................. ........................ ..................... 102

D.4 M-value, Failure Strain, Fracture Energy, DCSE, Initial Tangent Modulus,
D O an d D ..................................................................10 3

D.5 Dissipated Creep Strain Energy Calculation....... .................... .............. 104

E .1 Estim ated Loading Stresses (psi) ........................................ ....................... 106

E .2 N f to Initiation for D C SE ............................................... ............................ 107

E .3 N f to Initiation for F E ............................................................................. ..... .......107

E .4 N f to P ropagate 50m m ............................................................................. ...... 108

F.1 Summary of Traffic Loading (Annual ESALS in 1000)............... .....................110

F.2 Sum m ary of Loading Stress (psi)............................................... ........ ....... 110

F.3 Summary of Mixture Test Results Needed for Cracking Model............................111

F.4 Number of Cycles to Propagate Crack Length of 50mm ............... ...................111
















LIST OF FIGURES

Figure page

2.1 Fatigue Crack Growth Behavior (after Jacobs, 1995)................................................7

2.2 Dissipated Creep Strain Energy (after Zhang et al., 2001) ............. ..............8

3.1 Overview of Uncracked 1-75 Section............................................18

3.2 Overview of Cracked 1-75 Section ............................................... 18

3.3 O overview of SR -80 2U .................................................. ............................... 19

3.4 O overview of SR -80 C .................................................. ............................... 19

3.5 Longitudinal Crack from SR-80 C ...................................... ........ ............... 20

4.1 Cutting M machine .................. .............. .................. ........ .. ............ 26

4.2 ID T Testing D vice ................................................... .... .......... .... 27

4 .3 D ehum idifying C ham ber............................................................... .....................27

4.4 Tem perature Controlled Cham ber....................................... ......................... 28

4.5 Testing Sample with Extensometers Attached............... ............... ............... 28

5.1 Air Void Content and Comparison Between WP and BWP Sections ....... ........ 30

5.2 Effective Asphalt Content (%) ......... ........ .. ................. .................... 31

5.3 Gradation Curves for 1-75 1U and 1C ........................................... ............... 33

5.4 Gradation Curves for 1-75 2U and 3C............................... ...............34

5.5 Gradation Curves for SR-80 1C and 2U ...................................... ............... 35

5.6 Film Thickness ( m ) ..................................................... .... .. .......................36

5.7 B inder V iscosity (Poise)................................................. .............................. 37

5.8 R esilient M odulus (GPa) .......................................................... ............... 38









5.9 Creep Com pliance at 100 sec. (1/GPa) ....................................... ............... 39

5.10 Tensile Strength (M Pa) ................................................. ............................... 40

5.11 Failure Strain (m icrostrain) ............................................ .............................. 41

5 .12 m -v alu e ...................................... .................................................. 4 2

5.13 Fracture Energy D ensity (K J/m 3).......................................................... .... ......... 43

5.14 D C SE (K J/m 3) ......................................................... ................... 44

5.15 Loading stresses (psi) ............................................... .. ...... .. ............ 47

5.16 Number of Cycles to Failure for DCSE at 0 C ................................................. 49

5.17 Number of Cycles to Failure for DCSE at 10 C ............................................. 49

5.18 Number of Cycles to Failure for DCSE at 20 C............................................. 49

5.19 Number of Cycles to Failure for FE at 00 C....................................... ............... 50

5.20 Number of Cycles to Failure for FE at 100 C .................. .............................. 50

5.21 Number of Cycles to Failure for FE at 200 C .................. .............................. 50

5.22 Comparison between Nf of DCSE and FE limits at 00 C ........................................ 51

5.23 Comparison between Nf of DCSE and FE limits at 100 C ...................................... 51

5.24 Comparison between Nf of DCSE and FE limits at 100 C .......................................51

5.25 Number of Cycles to Failure to 50mm at 00 C..................................................... 52

5.26 Number of Cycles to Failure to 50mm at 10 C...................................... ........... 52

5.27 Number of Cycles to Failure to 50mm at 20 C...................................... ........... 52

6.1 Dimax for values of DCSE and m-value................................... ..............57

6.1 N to propagate 50m m ............................................ ..................................... 58

6.3 R relationship betw een a, St, and ....................................... ......................... 60

6.3 K HM A R atio for all sections............................................................ .....................62

6.4 KHMA Ratio for sections 0.75 KJ/m3 < DCSEf < 2.5 KJ/m3 ...................................63

6.5 Gradations of Low KHMA Ratio Sections with 12.5mm Nominal Aggregate Size ..65









6.6 Gradations of High KHMA Ratio Sections with 12.5mm Nominal Aggregate Size..66

6.7 Gradations of Low KHMA Ratio Sections with 9.5mm Nominal Aggregate Size ....67

6.8 Gradations of High KHMA Ratio Sections with 9.5mm Nominal Aggregate Size....68

6.9 Case 1 Gradation Comparison for 12.5mm Nominal mixes ..................................69

6.10 Case 2 Gradation Comparison for 12.5mm Nominal mixes ..................................69

6.11 Case 1 Gradation Comparison for 9.5mm Nominal mixes ..............................70

6.12 Case 2 Gradation Comparison for 9.5mm Nominal mixes ..............................70

6.13 Relationship Between Nf and Required KHMA Ratio................. ..... .............71

6.14 FT vs. Traffic loading (ESALS/year x 1000).......................... ...................72

6.15 Relationship Between Traffic Level and Fs .................................. .................73

6.16 Fs vs. Traffic (ESALS/year x 1000).......... ... ................................... 74

6.17 Minimum KHMA Ratio Required vs. Traffic Levels ...........................................75

A 1 D elections for 1-75 1U ................................................ ................................ 80

A .2 D elections for 1-75 C ................................................. ............................... 81

A .3 D elections for 1-75 2U ................................................. ............................... 83

A .4 D elections for 1-75 3C ................................................. ............................... 84

A .5 D elections for SR -80 2U ............................................... .............................. 86

A .6 D elections for SR -80 1C ...................................... .............................................87

A.7 Measured and Computed Deflections for 1-75 1U Location 03 ...............................88

A.8 Measured and Computed Deflections for 1-75 1U Location 06 .............................88

A.9 Measured and Computed Deflections for 1-75 1U Location 09 ..............................88

A. 10 Measured and Computed Deflections for 1-75 1C Location 03 .............................89

A. 11 Measured and Computed Deflections for 1-75 1C Location 08 ..............................89

A. 12 Measured and Computed Deflections for 1-75 1C Location 10............................89

A. 13 Measured and Computed Deflections for 1-75 2U Location 02.............................90









A. 14 Measured and Computed Deflections for 1-75 2U Location 09.............................90

A. 15 Measured and Computed Deflections for 1-75 2U Location 02.............................90

A. 16 Measured and Computed Deflections for 1-75 1C Location 03 .............................91

A. 17 Measured and Computed Deflections for 1-75 1C Location 04............................91

A. 18 Measured and Computed Deflections for 1-75 1C Location 06.............................91

A. 19 Measured and Computed Deflections for SR-80 2U Location 05 ...........................92

A.20 Measured and Computed Deflections for SR-80 2U Location 09 ........................92

A.21 Measured and Computed Deflections for SR-80 2U Location 01 ........................92

A.22 Measured and Computed Deflections for SR-80 1C Location 02..........................93

A.23 Measured and Computed Deflections for SR-80 1C Location 03..........................93

A.23 Measured and Computed Deflections for SR-80 1C Location 03..........................93

B.1 Cracking Ratings from 1-75 1U and C ....................................... ............... 95

B.2 Cracking Ratings from 1-75 2U and 3C ....................................... ............... 95

B .3 Cracking Ratings from SR-80 2U and 1C ...................................... .....................96

F.1 Effective Asphalt Content (% ) ...................................................................112

F .2 P percent A ir V oids (% ) .......................................................................... ....... .. 113

F.3 Theoretical Film Thickness (microns) ........... ................................ ............114

F.4 V M A (% )......................................................................... ........... 115















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering


DEVELOPMENT OF SPECIFICATION CRITERIA TO
MITIGATE TOP-DOWN CRACKING

By

Adam Paul Jajliardo

May 2003

Chair: Dr. Reynaldo Roque
Major Department: Civil and Coastal Engineering

One of the most common types of pavement distress in the State of Florida is

surface initiated longitudinal wheel path cracking, which is commonly refereed to as top-

down cracking. Prior research has indicated that top-down cracking is initiated by critical

tensile stresses in the surface of the asphalt pavement due to modern radial truck tires.

These cracks then propagate downward by the combined effects of temperature and

loads.

Several roadway sections were chosen for this study that exhibited top-down

cracking. Crack free sections were also studied that had similar structures and traffic as

the cracked sections. Asphalt Concrete cores were taken from each section for laboratory

testing to determine properties and characteristics of the mixture, binder, and aggregate.









Traffic volume, structure, and age data were collected for each section and Falling

Weight Deflectometer (FWD) tests were performed to determine the moduli of the

structural layers.

The mixture was tested using the Superpave Indirect Tensile Test (IDT). Resilient

modulus, creep compliance, and tensile strength tests were performed. The results of the

IDT test were used in the crack growth model developed at the University of Florida to

determine whether the number of cycles predicted to initiate and to propagate cracking in

the pavement correlated well with observed field cracking performance.

The analysis led to the identification of a mixture fracture toughness parameter

(KHMA(min)), which allows for the evaluation of mixture top-down cracking performance

by incorporating the affects of mixture properties and pavement structural characteristics.

The parameter did an excellent job indicating cracked from uncracked sections. It was

concluded that the parameter is suitable for the development of specification criteria to

mitigate top-down cracking and specific recommendations were made for its

implementation. The parameter (KHMA Ratio) was also defined that allows pavement

sections to be compared according to their cracking resistance.














CHAPTER 1
INTRODUCTION

1.1 Background

Surface-initiated longitudinal wheel path cracking is one of the most common types

of pavement distress in Florida today. Conventional load-induced cracking has been

commonly assumed to initiate at the bottom of the pavement and propagate upwards.

Cores and trenched sections taken from substandard sections have clearly illustrated the

phenomenon of top down cracking.

Myers (1997) identified the potential mechanisms of longitudinal surface initiated

wheel path cracking. She determined that this mode of distress is caused by a tensile

failure due to high stresses under the ribs of radial truck tires combined with thermal

stresses. Myers (2000) also found that surface initiated longitudinal cracking occurs

primarily under critical conditions. Therefore, existing design and evaluation methods

that consider only average conditions are not adequate in explaining this distress. These

approaches do not characterize the actual contact stresses or the discontinuities that exist

in the field. Myers (2000) also found that temperature gradients had a strong effect on

the development of stresses, which are also not considered in traditional fatigue

approaches.

Through the study of field cores, Garcia (2002) found that that there was no clear

relationship between any one mixture property and the mixture performance. However,

he found that the cracking model developed at the University of Florida appeared to

adequately explain the differences in mixture performance observed in the field.









Surface initiated longitudinal wheel path cracking has resulted in significant

rehabilitation costs. A better understanding of the mechanisms of surface cracking and

the key mixture properties and characteristics is necessary. The development of

specifications and design criteria will result in more crack resistant mixtures.

1.2 Objectives

The primary objectives of this research are summarized below:

* Evaluate field sections to identify the mixture properties and characteristics that
most strongly influence surface cracking performance.

* Develop a design specification for asphalt mixtures that would mitigate surface
cracking in pavements.

1.3 Scope

This study focuses on the analysis of the key mixture properties and characteristics

that affect surface initiated longitudinal wheel path cracking. To accomplish this, over

300 cores were extracted from six field sections throughout the state of Florida. Data was

also included from past studies and includes a total of twenty two sections.

1.4 Research Approach

The first step in this study was to conduct a literature review in order to understand

the different approaches to understanding fatigue in pavements. The different mixture

properties and characteristics were also evaluated to determine their influence in the

cracking performance of mixtures.

Next, the field sections were chosen from a number of potential sections through

visual inspection. Data from each of these sections was collected including the age,

structure, and traffic. Cracking performance over the life of the pavement was taken

from the FDOT Flexible Pavement Condition Survey Database.









Field cores were extracted from the roadway and each core was measured and cut

into test specimens. Mixture properties were determined using the Superpave Indirect

Tension Test (IDT). The specimens were broken down and the binder was extracted.

Binder and aggregate were obtained for evaluation.

Falling Weight Deflectometer (FWD) tests were performed on all of the sections in

order to determine the moduli of each pavement layer. The results were used to calculate

pavement stresses which were used to predict number of cycles to failure using the crack

growth model developed at the University of Florida.

The results of the crack growth model were analyzed in combination with results

from past studies at the University of Florida. The purpose of this was to evaluate the

effects of structure, environment, and mixture properties and characteristics on cracking

performance to develop specification criteria that would mitigate surface cracking in

asphalt pavements.














CHAPTER 2
LITERATURE REVIEW

A literature review was undertaken in order to understand the mixture

characteristics and properties that affect crack development and propagation. Several

different fatigue approaches were reviewed and their significance was determined when

discussing longitudinal surface-initiated top down cracking. It was also important to

review previous studies that investigated surface cracking in the field.

2.1 Fracture in Asphalt Pavements

Among all the types of failure in pavement, cracking is one of the most

predominant. Many factors influence cracking in pavement such as the pavement

structure and the mixture characteristics.

There are two main types of cracking in asphalt pavements. These are thermal

cracking and fatigue cracking. Thermal cracking is caused by the stresses that are

induced when low ambient temperatures cool the surface of the road. Fatigue cracking is

associated with traffic loading and is generated through repeated stresses. Myers (1997)

found that a probable cause of longitudinal surface initiated wheel path cracking is the

high tensile stresses caused by modern radial truck tires at the tire-pavement interface.

These stresses may be intensified by thermal stresses at the surface.

2.2 Mechanisms of Fracture in Asphalt Pavements

2.2.1 Traditional Fatigue Approach

The traditional fatigue approach is based on the assumption that the maximum

tensile strains are located at the bottom of the asphalt concrete layer. These strains









develop cracks and propagate from the bottom upward into the AC layer. Several fatigue

models have been developed to explain this phenomenon.

One of the first fatigue models was presented by Monismith et al. (1985). The

following relationship defines the fatigue behavior of a particular mixture:


Nf =AK K


where, Nf is the number of load applications to failure, A is a factor based on asphalt

content and degree of compaction, ;t is the tensile strain, Smix is the mixture stiffness and

a and b are constants determined from beam fatigue tests.

The Asphalt Institute developed the following empirical relationship in 1982 for a

standard mix with an asphalt volume of 11% and an air void volume of 5%:

Nf = .0796(,)3 291 (E *) 854

where, Nf is the number of load applications to cause fatigue cracking in 20% of the

pavement area, ;t is the tensile strain at the bottom of the surface layer, and E* is the

dynamic modulus of the asphalt mixture.

Another equation used to calculate the fatigue life of a mixture was developed

under the SHRP program (Sousa et al., 1996). As in the previous two equations for

fatigue life, it is a function of the mixture stiffness and asphalt content.

N, = S, *2.738x105 e00 77oBBSo-2720

where, Nf is the number of load cycles to failure, e is the base of the natural logarithm,

VFB is the voids filled with bitumen, Ft is the tensile strain, So is the loss of stiffness, and

Sf is a factor that converts laboratory measurements to anticipated field results. The

value of Sf is 10 for a pavement that is 10% cracked.









All of these models show that there are many variables that affect the fatigue

cracking performance of asphalt mixtures including mixture stiffness, AC content and air

voids. Also, this shows that there is no simple or reliable way to predict the fatigue life

of an asphalt mixture.

Myers (2000) found that the addition of a stiffness gradient in cracked asphalt

concrete significantly increased the tensile stresses in the surface of the AC layer. None

of the traditional fatigue approaches considers discontinuities (i.e. the presence of a

crack) in the asphalt layer or stiffness gradients in the asphalt layer that may be caused by

temperature or aging. The position of the load was also found to be a contributing factor.

Traditional approaches also do not allow for the possibility of changes in the load

positioning (wander) in the field. She concluded that current methods for the design and

evaluation are inadequate for longitudinal top-down cracking because they consider only

average conditions and this mechanism occurs primarily under critical conditions.

2.2.2 Fracture Mechanics Method

Another method to explain fracture in asphalt mixtures is the fracture mechanics

method, which introduces the concept of crack propagation. The rate of crack

propagation can be predicted using the following relationship known as "Paris Law":

da A(AK)n
dN

where a is the crack length, N is the number of load repetitions, A and n are parameters

depending on the mixture and AK is the difference between maximum and minimum

stress intensity factors during repeated loading. According to Ewalds and Wanhill

(1986), the fracture mechanics approach identifies three different stages. These are the

initiation phase where micro-cracks develop, the propagation phase where the micro-










cracks develop into macro-cracks and where crack growth becomes stable, and the

disintegration phase where the material fails, and crack growth is unstable.



Initiation Phase Propagation Phase Disintegration Phase




-z Fracture


0
-J

Threshold



Threshold LOG AK Critical
Figure 2.1. Fatigue Crack Growth Behavior (after Jacobs, 1995)

2.2.3 Dissipated Creep Strain Energy

Roque et al. (1997) found that the Dissipated Creep Strain Energy (DCSE) limit is

one of the most important factors that control crack performance in asphalt concrete

mixtures. The DCSE limit is the difference between the fracture energy (FE) and the

elastic energy (EE) at the instant of failure. The fracture energy is obtained from a

strength test as the area under the stress strain curve up to the point where the specimen

begins to fracture. The elastic energy can be obtained from the resilient modulus (MR).

Zhang (2000) introduced the concept of a threshold between micro-damage and

macro-cracking. Micro-damage was defined to be damage that was determined to be

healable. Macro-cracking was determined to be non-healable damage, even over long

rest periods and temperature increases. Zhang (2000) found that if the threshold was not

reached, cracks would not initiate and the mixture would be able to heal. Conversely, if

the threshold was reached the crack would grow and the mixture would not be able to









heal. She determined that the dissipated creep strain energy limit (DCSEf) was a suitable

threshold. Zhang introduced a fundamental crack growth law that is based on this work.

















S Strain g0 If

Figure 2.2. Dissipated Creep Strain Energy (after Zhang et al., 2001)

2.3 Mixture Properties Related to Fatigue Resistance

Many different material properties influence the fatigue resistance of asphalt

concrete mixtures. Therefore, it is necessary to review each of these properties to obtain

a clear understanding of fatigue resistance in asphalt pavements.

2.3.1 Mixture Stiffness

The mixture stiffness is defined as the ratio of the stress to the strain. For asphalt

mixtures, the stiffness is a function of time, temperature, and loading. The stiffness of an

asphalt mixture is affected by the binder stiffness, gradation, air void content, and asphalt

content. As a mixture ages the stiffness increases due to oxidation of the binder. This

increases the stiffness of the mixture and produces a mix that is more brittle and less

crack resistant.









2.3.2 Air Void Content

The amount of permeable air voids in a mix is related to the degree that the binder

is exposed to air and water. The exposure of binder to air and water results in the

oxidation of the binder and an increase in the rate of age hardening. The increase in age

hardening increases the stiffness and brittleness of a mixture.

The air void content is a function of aggregate gradation and degree of compaction.

Monismith et al. (1985) found that by increasing the air void content excessively resulted

in a decreased fatigue life.

2.3.3 Voids in the Mineral Aggregate (VMA)

VMA is the volume of the inter-granular void space between the aggregate particles

of a compacted pavement mixture. This void space includes the air voids and the asphalt

not absorbed into the aggregate. VMA is a function of degree of compaction, aggregate

gradation, aggregate shape, and air voids. It is an important factor in the durability of

asphalt mixtures. Generally, increased VMA values will increase the durability of a

mixture. Excessive VMA with high asphalt content will affect the durability adversely

because the high binder content tends to allow the aggregate particles to be pushed apart.

In the Superpave Mix Design Procedure (SHRP, 1993), minimum VMA is a

function of the nominal maximum aggregate size. Nukunya (2001) questioned the use of

the same VMA criteria for both fine and coarse mixes and recommended that

requirements should differ for the two.

2.3.4 Asphalt Content and Theoretical Film Thickness

Asphalt content is a very important factor in the cracking resistance of a mixture.

Asphalt content affects many material properties including air void content and film

thickness.









Lower asphalt content has been generally associated with inadequate amounts of

asphalt in a mixture. Monismith (1981) found that there is an upper limit to the amount

of asphalt that can be incorporated in a mixture, but that this limit should be approached

in order to increase the fatigue resistance. Pell and Taylor (1969) found that once the

optimum asphalt content is exceeded, there will be a decrease in fatigue resistance.

Valkering and Van Gooswilligen (1989) found that an approximate 1% decrease in the

binder content was found roughly to halve the traffic-related fatigue life.

The theoretical asphalt film thickness is a function of the effective asphalt content

and the surface area of the aggregate particles. For any given asphalt content, as the

surface area of the aggregate particles increases the theoretical asphalt film thickness

decreases. Very thin asphalt films contribute to excessive aging of the binder and in turn,

more brittle mixes and decreased cracking resistance. Thicker asphalt films contribute to

a more flexible and durable mixture. Kandhal and Chakraborty (1996) suggested a

minimum asphalt film thickness to produce durable mixtures. They concluded that an

optimum film thickness for HMA, compacted to 4 to 5% air void content, should be

higher than 9 to 10 microns.

2.3.5 Binder Viscosity

Pell and Taylor (1969) concluded that an increase in binder viscosity resulted in an

increase in fatigue resistance. Malan et al. (1989) concluded that higher viscosity

asphalts proved to be more crack resistant on lightly trafficked roads, while lower

viscosity asphalts resulted in better crack resistant mixtures on highly trafficked roads.

This can be explained by the constant kneading effect of the moving loads on high traffic

pavements. This kneading effect brings the volatiles to the surface of the pavement and

prevents excessive viscosity gradients.









The viscosity of an asphalt binder is influenced by aging and maybe more

importantly, by temperature. To prevent premature cracking, the binder viscosity is

chosen based on the climate of the region where the mixture will be placed. In low

temperature climates, unusually low viscosity binders should not be used because of the

risk of extreme temperature shrinkage.

2.3.6 Aggregate Gradation

Aggregate gradation plays a very important role in the structure of a mixture. The

quality of aggregate interlock is primarily responsible for the mixture's response to load.

The aggregate gradation affects VMA and asphalt film thickness.

The opinions on the effect of gradation on fatigue resistance are divided.

Monismith et al. (1985) found there is an insignificant effect on fatigue resistance that is

not explained by air void content and asphalt content. Malan et al. (1989) concluded that

continuously graded asphalt mixture designs are less susceptible to surface cracking than

gap graded and semi-gap-graded designs. Continuously-graded mixtures tend to have

higher asphalt film thickness and are more able to dissipate the shrinkage stresses.

2.4 Previous Studies

Sedwick (1998) conducted a study on top-down longitudinal wheel path cracking

that examined cores taken from the field. He used these cores to identify the mixture

properties and characteristics that would lead to the development of surface cracking. He

determined that fracture energy density was a good indicator of cracking performance

when other conditions such as pavement structure, traffic, and environmental effects are

the same. He also found that samples from the field with fracture energy densities lower

that 1.0KJ/m3 indicated a poor crack resistant mixture.









Garcia (2002) also conducted a study on longitudinal wheel path cracking using

field cores. He used these cores to identify the key factors that contribute to the

development of surface cracking. He determined that there was no clear relationship

between any single material property that would adequately describe the cracking

performance. However, the results of the HMA fracture mechanics based model

developed at The University of Florida appeared to properly explain the difference in

mixture performance. He also determined that the effects of the pavement structure and

thermal stresses were significant when comparing the relative cracking performance of

pavement sections.

2.5 Summary

* Traditional fatigue approaches do not adequately explain the phenomenon of
longitudinal wheel path cracking.

* Fracture mechanics provides a solid foundation for understanding cracking in
asphalt pavements. It introduces the concept of initiation, propagation, and
disintegration.

* Dissipated Creep Strain Energy is one of the most important factors when
considering the cracking performance of asphalt mixtures. The Dissipated Creep
Strain Energy limit (DCSEf) can be used as a threshold between micro-damage and
macro cracking.

* Mixture stiffness is a function of temperature, time, and loading. Excessively stiff
mixtures are generally less crack resistant.

* Permeable air voids affect the degree of age hardening. Excessively high air voids
will decrease the crack resistance of a material.

* Film thickness is a function of gradation and asphalt content. Thicker film
thickness results in a mixture that is more durable, flexible, and crack resistant.

* Aggregate gradation plays a defining role in the structure of a mixture. Mixtures
that are more continuously graded are more crack resistant.

* In previous studies, it was found that there was no clear relationship between any
one-mixture property and cracking performance. All of the material properties, as






13


well as the pavement structure must be examined in order to describe a mixture's
cracking performance.

* The crack growth model developed at The University of Florida appears to
adequately represent the cracking mechanisms of asphalt mixtures in the field.
















CHAPTER 3
DESCRIPTION OF TEST SECTIONS

Six sections from three locations were chosen for this study. These sections were

chosen in pairs of good and poor performance with similar structure, loading, and age,

but with different mixtures. This chapter provides a description of the sections.

3.1 Locations and Age

The six sections were all extracted from locations in southwest Florida. Sections

1U and 1C were taken from 175 in Charlotte County. Sections 2U and 3C were taken

from 175 in Lee County. SR 80 sections were also taken from Lee County. Table 3.1

summarizes the locations of the sections.

Table 3.1. Location of the Sections
Section Section Name Condition Code County Section State Mile
Number Limits Posts

1 Interstate 75 U I75-1U Charlotte MP 149.3 MP 161.1 0-11.8
Section 1
2 Interstate 75 C I75-1C Charlotte MP 161.1 MP 171.3 11.8 22.0
Section 1
3 Interstate 75 U I75-2U Lee MP 115.1 MP 131.5 0- 16.4
Section 2
4 Interstate 75 C I75-3C Lee MP 131.5 MP 149.3 16.4 34.1
Section 3
5 State Road 80 C SR 80-2C Lee From East of CR 80A 10.8 13.6
Section 1 To West of Hickey Creek Bridge
6 State Road 80 U SR 80-1U Lee From Hickey Creek Bridge 13.6 18.3
Section 2 To East of Joel Blvd.

The age of the sections is defined as the time from the most recent resurfacing. The

age of each section is summarized in Table 3.2.









Table 3.2. Age of the Sections
Section Year Let Age as of 2003

I75-1U 1989 14
I75-1C 1988 15
I75-2U 1989 14
I75-3C 1988 15
SR 80-2U 1984 19
SR 80-1C 1987 16

3.2 Pavement Structure

The layer moduli were determined with the Falling Weight Deflectometer (FWD).

The values were then back calculated using elastic layer analysis. The FWD procedure

used the standard SHRP configuration for the sensors (i.e. 8", 12", 18", 24", 36", and

60"). For each section, ten tests were conducted in the travel lane in the wheel path at

relatively undamaged locations, on both sides of the coring area. A half-inch hole was

drilled in the pavement and filled with mineral oil or glycol for heat transfer and the

pavement temperatures were recorded. The pavement surface and ambient temperatures

were also recorded. A 9-kip seating load was applied, followed by 7, 9, and 1 Ikip loads.

Deflection measurements at each of the sensors were recorded. The layer thickness and

back-calculated moduli appear in Tables 3.3 and 3.4, respectively. The base and sub-

base thickness were not available so a typical thickness of 12 inches was assumed for the

back calculation analysis.

Table 3.3. Thickness of the layers (in)
Section Friction Course AC Base Sub-base
I75-1U 0.44 6.23 12 12
I75-1C 0.51 6.54 12 12
I75-2U 0.46 7.42 12 12
I75-3C 0.62 6.47 12 12
SR 80-2U 0.80 6.29 12 12
SR 80-1C 0.37 3.38 12 12










Table 3.4. Layer Moduli for each Section (ksi)
Section AC Base Sub-base Sub-grade
I75-1U 1000 64 51 36
I75-1C 800 55 50 30
I75-2U 1000 107 90 31
I75-3C 900 60 35 36
SR80-2U 500 57 46 19
SR80-1C 800 44 61 28

3.3 Traffic Volume

The traffic volumes for each section are shown in table 3.5. These values are

expressed in thousands of ESALS. The traffic volumes vary from 207 K for section SR-

80 2U to 674 K for section 1-75 3C.

Table 3.5. Traffic Volumes for each Section (Millions)
Section Traffic (ESALS/year x 1000)
I75-1U 558
I75-1C 573
I75-2U 576
I75-3C 674
SR80-2U 207
SR80-1C 221

3.4 Environmental Conditions

The environmental conditions were similar for all the test sections. Florida has a

humid climate with average yearly temperatures between 200 and 250 C. Pavement

temperatures during the summer months can increase considerably.

3.5 Performance of the Sections

3.5.1 Overview

The Flexible Pavement Condition Survey Database is a record maintained by the

Florida Department of Transportation. This record contains ratings on the ride, rutting,

and cracking performance of every pavement section supported by the FDOT. The

primary purpose of this record is to prioritize pavement sections for rehabilitation.









The ratings for cracking are based on crack width and cracked surface area. Values

between 0 and 10 are given to a pavement section depending on the size of the crack

widths and the surface area of the pavement that is cracked. The value of 10 is given to a

pavement that is crack free. Appendix B contains the crack ratings of all the sections.

Unfortunately, this rating judges only the appearance of the surface and gives no

indication of the actual depth of the cracks. For example, a pavement with a high amount

of cracking in the friction course would be given a low crack rating even if the cracks did

not propagate further into the pavement. Also, a pavement with a small number of cracks

would be given a high rating even though the cracks may extend well into the pavement.

Therefore, to gauge the actual extent of the cracking, it was necessary to core the

pavement directly though the crack and measure the crack depths manually.

3.5.2 Field Observations

Before the coring was performed, a field trip was taken to each section to observe

and take pictures. Figure 3.1 shows a typical uncracked section for 1-75. Figure 3.2

shows a cracked section. The uncracked section appears to be in an acceptable condition.

The cracked section exhibits a moderate amount of cracking as well as wheel rim

markings. The cracks appear in and to the side of the wheel paths in the travel lane.

Figures 3.3 and 3.4 show the uncracked (SR80 2U) and cracked (SR80 1C) sections

respectively. The uncracked section appears to be in a very acceptable condition with a

surface free from cracks. The cracked section is heavily cracked with continuous cracks

appearing in the wheel paths of both lanes. Figure 3.5 shows a close-up of a crack.























;| .. t .


i- 1i. o.

Fi";g r 3 1'v -- .e" o Unrc 7. .


Figure 3.1. Overview of Uncracked 1-75


''
...,
_,t


-r
Se. cti



5 Section


Figure 3.2. Overview of Cracked 1-75 Section


























































Figure 3.3. Overview of SR-80 2U


.M ;g
e-r.r


Figure 3.4. Overview of SR-80 1C


~

-'


3

.r






























Figure 3.5. Longitudinal Crack from SR-80 1C














CHAPTER 4
MATERIALS AND METHODS

After the cores were extracted from the roadway, they were measured and saw-cut

into individual testing specimens. The bulk specific gravity of each specimen was

measured. The specimens were then tested using the Superpave indirect tensile test

(IDT) developed by Roque et al. (1997). One specimen was used to determine the

Maximum Theoretical Density using the Rice test. The binder and the aggregate were

separated from representative specimens from each section. These were used for further

testing.

4.1 Extraction of the Field Cores

Several cores were taken from each of the six sections. Cores were extracted from

the wheel path as well as between the wheel paths. Cores were also taken through the

crack in order to measure the actual crack depths. The cores were marked for traffic

direction. This is necessary because the failure that was observed in the field is a tensile

failure perpendicular to the direction of traffic. It was important to keep the direction

consistent when performing the IDT tests. A total of 46 cores were extracted from each

section. Eighteen cores each were taken from the wheel path and eighteen from between

the wheel paths. Ten cores were taken through the cracks for each section. All of the

cores were extracted using a truck mounted coring rig. The truck-mounted rig was used

to minimize damage that may occur to the samples during the coring process.









4.2 Measuring and Cutting the Field Cores

Upon inspection in the laboratory, the thickness of each lift was measured and

recorded. The crack depths were also measured and recorded. Since the cracks originate

at the surface, the layer immediately beneath the friction surface is primarily responsible

for crack initiation and propagation. This layer was chosen for the purposes of this study.

This layer was identified for each core and marked for cutting. Figure 4.1 shows a

picture of the machine used to cut the samples.

The thickness of the sample used for the IDT testing is typically between 1 to 2

inches. The actual thickness of the specimens varied from 1 to 1.81 inches, depending on

the thickness of the layer as well as the quality of the cores. The average thickness of the

samples for each section is shown in Table 4.1.

After the specimens were cut, they were marked for future identification. Since the

cutting process involves water, the specimens were placed in an air-conditioned

environment for several days until their natural moisture content was reached. The bulk

specific gravity of each specimen was measured.

4.3 Selecting Samples for Testing

Nine specemins were needed from each section in order to test the mixture at three

temperatures. The samples were chosen by selecting nine samples having a bulk specific

gravity (Gmb) closest to the average Gmb of the section. Table 4.2 shows the average bulk

specific gravities for each section.









Table 4.1. Average thickness
Section Thickness
(in.)
I75-1U 1.17
I75-1C 1.10
I75-2U 1.00
I75-3C 1.06
SR 80-2U 1.31
SR 80-1C 1.81

Table 4.2. Average Bulk Specific Gravity
Average Gmb
Section WP BWP
I75-1U 2.302 2.241
I75-1C 1.860 2.274
I75-2U 2.284 2.221
I75-3C 2.281 2.208
SR 80-2U 2.232 2.181
SR 80-1C 2.267 2.204
Note:
WP: Wheel Path
BWP: Between Wheel Path

4.4 Crack Rating

The cracked cores were taken and the crack depths were measured and recorded.

Sedwick (1998) defined a crack rating criteria based on the average crack depth measured

for a given section. This criteria assigns a value between 0 and 10 based on the length of

the measured crack depths. Table 4.3 shows the rating criteria used by Sedwick. Table

4.4 shows the average crack depths for the six sections and their corresponding

performance rating. Cracking was found to be especially severe in section SR-80 1C

where in some cases the cracks extended completely through the cores.









Table 4.3. Cracking Criteria (After Sedwick, 1998)
Crack Depth (in) Rating
< 0.25 10
0.26 0.75 8
0.76- 1.25 6
1.26-2.00 4
2.01-3.00 2
> 3.00 0

Table 4.4. Crack Ratings
Section Average Cracking Performance
Depth (in) Rating
175-1U Uncracked 10
I75-1C 2.21 2
175-2U Uncracked 10
I75-3C 2.32 2
SR 80-2U Uncracked 10
SR 80-1C 2.56* 2
*Some samples cracked completely through

4.5 Mixture Testing

The testing procedures used in this study were developed for the FDOT by Roque

et al, (1997). The following tests were performed: Resilient Modulus, Creep

Compliance, and Tensile Strength. These tests were performed at three temperatures:

0C, 10C, and 20C. The results from the three tests were analyzed using software

developed at the University of Florida. This provided Resilient Modulus (GPa), Creep

compliance as a function of time (1/GPa), Tensile Strength (MPa), Failure strain

(microstrain), Fracture Energy (KJ/m3), m-value (the slope of the linear portion of the

creep compliance-time curve), and the Dissipated Creep Strain Energy to failure.

Poisson's ratio is also calculated for each of the three tests.

A gage placement jig was used to mount four aluminum gage points on each face

of the testing specimens. The samples were then placed in a relatively low humidity









chamber for forty-eight hours to eliminate any moisture that would affect testing. Before

testing, the samples were placed in the temperature-controlled chamber overnight to

stabilize the temperature. Before testing, the samples were fitted with knife edged gage

mounting blocks. A set of spring-loaded extensometers was placed on the knife-edges to

measure deformations. Figures 4.2 through 4.5 show pictures of the IDT testing

machine, the dehumidifying chamber, the temperature-controlled chamber, and a sample

with the extensometers attached.

4.6 Asphalt Extractions and Binder Testing

Two samples from each section were broken down to determine the Theoretical

Maximum Specific Gravity (Gm) or Rice Gravity according to AASHTO T 209-94. The

Gmm value for each section made it possible to calculate the air void content for each

sample. These samples were then placed in an asphalt extraction device that uses

trichloroethylene (TCE) to separate the binder from the aggregate. The binder-TCE

mixture was then placed in an extraction device, which evaporated the TCE. Viscosity

tests were performed on the binder at 60C using the Brookfield Thermosel Apparatus.

These tests were performed according to ASTM D 4402-87.

4.7 Aggregate Tests

Once the aggregate was separated from the binder, it was placed in an oven

overnight to evaporate any remaining TCE. Following this, a washed sieve analysis was

performed according to ASTM C 117 to determine the amount of material passing the No.

200 sieve. A dry sieve analysis was then performed to determine the gradation of each

mixture following ASTM C136. The specific gravity and absorption of the fine and

coarse aggregates was performed according to ASTM C128 and ASTM C127

respectively.









4.8 Volumetric Properties

Using the results of the binder and aggregate tests, several volumetric properties

were calculated. These were the effective asphalt content, the Voids in the Mineral

Aggregate (VMA), and the theoretical film thickness. The effective asphalt content was

calculated using the results from the extraction-recovery process as well as the percent

absorption. The VMA was calculated using the bulk specific gravity of the mixture, the

specific gravity of the aggregate and the aggregate content. The theoretical film

thickness was obtained from the Hveem method. The Hveem method calculates the film

thickness by approximating the surface of the aggregate using surface area factors. These

surface area factors are multiplied by the percentage passing for each sieve. The film

thickness is calculated by dividing this surface area by the volume of effective asphalt.

The calculations of these volumetric properties are summarized in Appendix C.


Figure 4.1. Cutting Machine






























Figure 4.2. IDT Testing Device


Figure 4.3. Dehumidi















Figure 4.4. Temperature Controlled Chamber


Figure 4.5. Testing Sample with Extensometers Attached


,--l














CHAPTER 5
ANALYSIS AND FINDINGS

This chapter presents the results of the mixture and binder results as well as the

analysis of the Falling Weight Deflectometer Test (FWD). This chapter also provides

analysis as to the material properties that were related to the cracking performance of the

mixtures. This analysis was conducted through the comparison of sections with similar

age, pavement structure, and traffic conditions.

5.1 Volumetric Properties and Extraction-Recovery Results

The cores that were obtained in the field were cut and specimens were used to

determine the bulk specific gravity of the mixture. Samples from each section were

broken down and their Theoretical Maximum Specific Gravity was determined. These

samples were also put through an extraction and recovery process to determine the

asphalt content and viscosity as well as several aggregate properties. From these results,

several volumetric properties were calculated including effective asphalt content, VMA,

and theoretical film thickness. Results are shown in the sections that follow.

5.11 Air Void Content

The air void content for each section was calculated using the bulk specific gravity

of the mixture (Gsb) and the theoretical maximum specific gravity (Gm). The average

void content and standard deviation for each of the sections is shown in Table 5.1. The

comparisons between the sections are shown in Figure 5.1.









Table 5.1. Air void content
WP BWP
S Air Void Standard Air Void Standard
Section
Content Deviation Content Deviation
I75-1U 1.86 0.13 3.21 0.35
I75-1C 2.88 0.18 5.39 0.52
I75-2U 3.72 0.57 6.93 0.82
I75-3C 4.37 0.51 7.24 0.88
SR 80-2U 4.72 1.10 7.53 0.69
SR 80-1C 2.77 0.55 5.71 0.53


8.00
7.00
6 6.00
5.00
.WP
: 4.00
C BWP
3.00
2.00
1.00
0.00
I75-1U I75-1C I75-2U I75-3C SR 80-2U SR 80-1C

Section

Figure 5.1. Air Void Content and Comparison Between WP and BWP Sections.

For all the sections, there seemed to be a similar difference in air void content

between the WP cores and the BWP cores. The 1-75 sections 1U and 1C showed a

difference in air voids with values ranging between 2% and 3% for the WP cores and 3%

and 5.5% for the BWP cores. The 1-75 sections 2U and 3C exhibited approximately the

same percentage of air voids with approximately 4% for the WP cores and 7% for the

BWP cores. The SR-80 sections 2U and 1C also showed a difference in air voids. The

values ranged from approximately 5% for section 2U to 3% for section 1C for the WP









cores and from 7.5% for 2U and 5% for 1C for the BWP cores. There was no clear

relationship between the air void content and the cracking performance.

5.1.2 Effective Asphalt Content

The effective asphalt content was determined from the percent asphalt absorbed

using the aggregates from the extraction-recovery process. Low asphalt content means

poor coating of the aggregate particles and poor fatigue resistance. The FDOT requires

effective asphalt contents to be greater than 5%. The figure below shows that none of

these sections met the requirement. There was little difference between the cracked and

uncracked pairs and no visible relationship between effective asphalt content and

cracking performance.


6.00

5.00

4.00

3.00

2.00

1.00

0.00
I75-1U I75-1C I75-2U I75-3C SR-80 2U SR-80 1C
Section

Figure 5.2 Effective Asphalt Content (%)

5.1.3 Aggregate Gradation

The aggregate gradation for each of the sections is shown in figures 5.3 to 5.5. The

gradations shown below are expected to be finer than the in place gradations. This is

because the aggregates used to compute the distributions were taken from samples cut









from cores and the sizes of the specimens were less than the 2500 grams required for a

gradation test. The resulting gradations will be finer than the actual gradations in the

field but since the size of the specimens were approximately the same it was assumed that

the shift in gradation due to these two effects was the same for all samples.

The aggregates from 1-75 sections 1U and 1C have a similar aggregate gradation

distribution. Each section has low dust contents and a significant amount of material

between the #100 and #8 sieves.

1-75 sections 2U and 3C are compared in Figure 5.3. The grain size distributions

for the two sections are also similar. Each has a low dust content, a significant amount of

material between the #100 and #30 sieves and a gap between the #30 and #8 sieves.

Section 3C appears to be a slightly more gap-graded mixture. Previous studies (Sedwick,

1998) found that gap graded mixtures are generally less crack resistant than more

continuously-graded mixtures. Section 3C appears to be slightly finer than section 2U.

SR-80 sections 2U and 1C are compared in Figure 5.4. Their distribution curves

are similar in that they both have a large amount of material between the #200 and #50

sieves and both have low dust contents. Section 2U has a finer distribution than section

1C. A finer gradation has also been linked to poor cracking performance because of the

greater tendency for the mixtures to have low asphalt film thickness. However, there

does not appear to be a strong relationship between the relative fineness of the mixture

and the cracking performance.



















100


90


80


70


60


50


40


30
SI75-1C

20 I75-1U
S- Max Density
10 Line


0
0 200 100 50 30 16 8 4 3/8 1/2 3/4 1
Sieve sizes


Figure 5.3 Gradation Curves for 1-75 1U and 1C



















100


90


80


70


60


50


40


30
20 0-0-I75-3C

20 I75-2U
Max Density
0 Line
10


0
0 200 100 50 30 16 8 4 3/8 1/2 3/4 1
Sieve sizes


Figure 5.4 Gradation Curves for 1-75 2U and 3C



















I UU


90


80


70


60


50


40


30
SR80-1C

20 SR80-2U
SMax Density
10 tLine
10



0 200 100 50 30 16 8 4 3/8 1/2 3/4 1
Sieve sizes


Figure 5.5 Gradation Curves for SR-80 1C and 2U










5.1.4 Theoretical Film Thickness

The theoretical film thickness was calculated using the Hveem method. The

Hveem method calculates the film thickness from the aggregate gradation and the asphalt

content. As previously discussed in Chapter 2, Kandhal proposed a minimum film

thickness of 9 to 10tm. From the data below in Figure 5.6, all of the sections have

theoretical film thickness of less than 9[tm and are generally below 6[tm, which is

considered excessively low. This may be due to the excessively low effective asphalt

contents. For the 1-75 1U and 1C and the SR-80 sections the uncracked sections had

higher film thickness than the cracked sections, while the reverse was found for the 175

2U and 3C sections.


5.80

5.60

5.40

5.20

5.00

4.80

4.60

4.40
I75-1U I75-1C I75-2U I75-3C SR-80 2U SR-80 1C
Section


Figure 5.6 Film Thickness ([tm)











100000
90000
S80000-
70000
& 60000
S50000 -
.W5 BWP
40000
30000
20000
10000
0
175-1U 175-1C 175-2U 175-3C SR 80- SR 80-
2U 1C
Section

Figure 5.7 Binder Viscosity (Poise)

5.1.5 Binder Viscosity

Figure 5.7 shows a comparison of the binder viscosities between the sections.

The binder viscosity values follow a similar trend with the air void content. Sections

with higher air void contents also exhibit higher binder viscosities. This can be explained

by the age hardening that occurs due to oxidation. A higher air void content generally

allows for an increased amount of oxidation that results in a higher rate of age hardening.

The SR80 sections show much higher binder viscosities than other sections. This may be

partially due to the older age of the pavement sections.

5.2 Mixture Results

The mixture test performed were resilient modulus, creep compliance at 100

seconds and tensile strength. These tests were performed at 00 C, 100 C, and 200 C. The

mixture properties that were obtained from these tests were the resilient modulus, creep

compliance, m-value, tensile strength, fracture energy density, failure strain, initial









tangent modulus, and the dissipated creep strain energy limit (DCSEf). The following is

a summary of the test results and a detailed analysis of each mixture property and how it

relates to mixture cracking performance.

5.2.1 Resilient Modulus

The resilient modulus (MR) is a measure of a material's elastic stiffness. This is a

function of the binder stiffness and the degree of aggregate interlock. Figure 5.8 shows

the values of MR for each of the sections at each of the three test temperatures.


S00 10 D 200

20

16 -
14
12
10 -
8-

6-
4 4
2
0
I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C
Section

Figure 5.8 Resilient Modulus (GPa)

The 1-75 1U and 1C sections exhibited almost the same MR values for 100 and 20.

Section 1C had slightly higher MR values at 00 with values only 13% over those of

Section 1U. 1-75 sections 2U and 3C also had similar values of MR. At 200, the values

were almost equal, at 100 section 3C was greater than 2U by 12.5% and at 00 section 2U

was greater than 3C by 10%. The results for the SR80 sections are similar to those of the

1-75 sections. The resilient modulus values for SR-80 2U and 1C are also close to equal.









At 00 and 100, the results for the two sections are almost identical. At 200 the value for

section 2U is higher than 1C by 20%. The data suggests that there is no clear relationship

between resilient modulus and cracking performance. However, tensile stresses will be

greater in the sections that have slightly greater resilient modulus values.

5.2.2 Creep Compliance

Creep compliance is related to the ability of a mixture to relax stresses especially

thermal stresses. Mixtures with higher creep compliances can relax stresses more quickly

than mixes with low creep compliances. Figure 5.9 shows the creep compliances at 100

seconds for each test section. For the 1-75 1U and 1C sections, the compliance values

were slightly higher for section 1U than those for 1C. In contrast, the compliance values

for 1-75 2U and 3C were very similar. For sections SR 80 2U and 1C, the uncracked

section had lower compliance values than the cracked section. There appears to be no

relationship between creep compliance and the cracking performance of the sections.


S00 100 0 200

3.500

S 3.000
2.500

2.000
S 1.500

1.000
0.500

0.000
I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C
Section


Figure 5.9 Creep Compliance at 100 sec. (1/GPa)









5.2.3 Tensile Strength

Tensile strength is the maximum tensile stress that the mixture can withstand

before failure. The indirect tensile strengths for each section at all three temperatures are

shown in Figure 5.10. The uncracked sections possessed slightly higher tensile strengths

than the cracked sections.


S0 100 20

3.50
S3.00

Figure 5.10 Tensile Strength (MPa)
2.00
S1.50 -
S1.00 -
S0.50 -
0.00
I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C
Section

Figure 5.10 Tensile Strength (MPa)

5.2.4 Failure Strain

Failure strain is the horizontal strain that is measured during the indirect tensile

strength test when cracking occurs. Failure strain is a direct measurement of the

brittleness of a mixture. In general, mixtures with high failure strains are more crack

resistant. The failure strains for each section at each temperature are shown in Figure

5.11. For all sections, the uncracked sections displayed higher failure strains than the

cracked sections excluding 1-75 sections 2U and 3C at 200 C, which were close to equal.










H0 10 ] 200

S3000

1 2500

2000

1500

S1000

500

0 -- -- --
I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C

Section


Figure 5.11 Failure Strain (microstrain)

5.2.5 m-value

The m-value is defined as the slope of the linear portion of the log creep

compliance log time curve. It is calculated by fitting the following relationship to the

creep compliance data:

D(t)= D0 +Dtm

where, D(t) is compliance at time t, Do and D1 are model parameters, and m is the m-

value. The m-value is an indirect measurement of the creep rate of a mixture. A mixture

with a higher m-value has a higher creep rate, which implies a higher rate of damage for a

given stress. However, it also means a higher rate of stress relaxation. Also higher m-

values are typically associated with softer binders and mixtures with higher Fracture

Energy thresholds.

Figure 5.12 shows the m-values for each section at all three temperatures. The

most significant difference in m-values between paired sections was found on SR-80.









The m-value for section SR-80 1C was higher than section 2U for all temperatures.

These values agree with the lower binder viscosity for SR-80 1C.


*00 100 20

0.6

0.5

0.4

0.3

0.2

0.1

0.0
I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C
Section

Figure 5.12 m-value

5.2.6 Fracture Energy Density and Dissipated Creep Strain Energy

Fracture energy density is defined as the energy per unit volume that is required to

fracture an asphalt mixture. It is calculated from the indirect tensile strength test by

computing the area under the stress-strain curve up to the point the sample starts to fail.

Previous studies (Sedwick, 1997) have determined that fracture energy is a reliable

indicator of the crack resistance of a mixture when other conditions such as pavement

structure and traffic are similar. He suggested that mixtures with fracture energy

densities of less that 1 KJ/m3 at 00 or 100 performed poorly in the field. Garcia (2002)

found that the pavement structure and thermal stresses were also important when

comparing the relative performance of asphalt mixtures.

The fracture energy densities are shown below in Figure 5.13 for all the sections at

the three temperatures. As the data indicates, the fracture energy densities for the









uncracked sections are greater than their paired cracked sections with the exception of the

1-75 2U and 3C sections at 200. The values of fracture energy for all cracked sections

were below 1 KJ/m3. All uncracked sections had fracture energy values equal to or

greater than 1 KJ/m3 at 100C. The fracture energy of the SR-80 uncracked section at 0C

was less than 1 KJ/m3. The much smaller traffic levels experienced by this section may

explain it's good performance in the field.


*00 100 20

3.0 -

2.5

2.0

1.5 -



S0.5

0.0
I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C
Section

Figure 5.13 Fracture Energy Density (KJ/m3)

The dissipated creep strain energy at failure (DCSEf) is defined as the fracture

energy minus the elastic energy (Zhang, 2000). The values of DCSEf for each section are

shown below in Figure 5.14 at the three test temperatures. Since DCSE is a function of

the fracture energy, it is reasonable that the results would display a similar trend. From

the results of the data, it can be seen that the uncracked sections posses higher DCSE

values than their respective cracked pairs except for sections 1-75 2U and 3C at 200.










00 100 E 200

3.0
2.5
2.0
1.5
S1.0
0.5
0.0
I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C
Section

Figure 5.14 DCSE (KJ/m3)

5.3 Non-Destructive Testing (FWD)

Falling Weight Deflectometer (FWD) testing was performed on each of the section

in order to determine the layer moduli in the pavement system. Back calculation analysis

was used to interpret the testing data. These modulus values were used including

information about the layer thickness to calculate the stresses at the bottom of the AC

layer. The measured deflections for each sensor and location are shown in Appendix A.

5.3.1 Pavement Structures

FWD testing was run at three deflection levels: high, intermediate, and low. The

deflection basin data was used to back calculate the moduli of each layer in the pavement

system using BISDEF at three locations along the length of each section. Table 5.2 shows

a summary of the results.









Table 5.2 Layer Moduli from FWD Analysis
Section AC Base Sub-base Sub-grade
I75-1U 1000 64 51 36
I75-1C 800 55 50 30
I75-2U 1000 107 90 31
I75-3C 900 60 35 36
SR80-2U 500 57 46 19
SR80-1C 800 44 61 28

The results of the back calculation show that there are some differences in the

structure of the pavement sections. 1-75 sections 1U and 1C have similar base and sub-

base moduli although the base modulus for 1U is slightly higher than 1C. One of the

largest differences is in the base and sub-base moduli for 1-75 sections 2U and 3C. The

base and sub-base moduli for section 2U are almost double those of section 3C. This had

a significant impact upon lowering the tensile stresses in section 2U. The structures for

the SR-80 sections were similar although the base stiffness for section 2U was slightly

greater that that of 1C. The sub-base stiffness is higher for section 1C but this may be

attributed to fitting error in the back calculation analysis and does not have a significant

effect on the pavement stresses.

The ratio of the asphalt concrete modulus to the base modulus (E1/E2) is a good

indicator of the bending stresses in the AC layer. A larger E1/E2 ratio generally indicates

higher bending stresses. Table 5.3 shows the E1/E2 ratios for each section. For each

paired cracked and uncracked section, the E1/E2 ratios were larger for the cracked

section. For the 1-75 sections 2U and 3C, the ratios were doubled for section 3C.









Table 5.3 E1/E2 Ratios
Section Temp E1/E2
0 34
I75-1U 10 25
20 17
0 45
I75-1C 10 29
20 21
0 25
I75-2U 10 14
20 10
0 41
I75-3C 10 28
20 20
0 46
SR80-2U 10 34
20 28
0 62
SR80-1C 10 45
20 30

5.3.2 Loading Stresses

Using the layer thickness and the layer moduli calculated from the FWD data, the

loading stresses were calculated at the bottom of the AC layer with BISAR. These

stresses were calculated at three loading levels: 7,000 lbs., 9,000 lbs., and 11,000 lbs. and

at the three testing temperatures of 00, 100, and 200. A complete summary of the loading

stresses appears in Appendix E. The modulus values used in the stress calculation were

the MR values calculated in the laboratory. Figure 5.4 shows the stresses for a 9,000-lb

load at all three temperatures. For any given temperature and load, the estimated stresses

in the bottom of the AC layer were greater for the cracked sections than for the uncracked

sections. The stresses for 1-75 sections 1U and 1C are almost identical and indicate

similar structural characteristics. The other sections displayed much greater differences

the their stresses.










1-75 section 3C had approximately 60% greater stresses than section 2U. The stresses for

SR-80 section 1C were in some cases double those of section 2U.


0 0l 10 020

450
400
350
300
4 250
S200
S150
100
50

I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C
Section

Figure 5.15. Loading stresses (psi)

5.4 Crack Growth Model

As mentioned earlier in Chapter 2, Zhang (2001) developed a fracture mechanics

based crack growth model at The University of Florida. This model was used to

determine the relative cracking performance of each roadway section.

The crack growth model explains two phases of crack development: initiation and

propagation. The initiation phase predicts the number of cycles of loading required to

exceed a threshold point and begin macro cracks. The propagation phase predicts the

number of cycles that are required to grow a crack a given distance starting with an initial

user-defined crack length. The stresses that were used as inputs for both phases of the

cracking model were calculated using BISAR at the bottom of the asphalt layer. Since

the stresses were calculated at the bottom and not at the top of the asphalt layer, the

results of the initiation and propagation phases of the cracking model can only be used to

compare the cracking performance in a relative manner.









For the initiation phase, the number of cycles required to achieve macro cracking -

were calculated for low, medium, and high loading levels (7000, 9000, and 11000 Lbs.)

at the three test temperatures of 00, 100, and 200 C. The model calculates the number of

cycles for both dissipated creep strain energy and fracture energy limits. The results of

the initiation phase appear in Figures 5.16 through 5.21.

For each temperature and for each load, it was found that the uncracked sections

required more cycles to reach their thresholds and initiate a crack than their paired

uncracked sections. In all cases, there were significant differences in Nf. 1-75 section 1U

required almost two times the Nf of section 1C to reach its threshold. 1-75 section 2U

required three times the Nf of section 3C to reach its threshold and SR-80 section 2U

required almost six times the Nf of SR-80 section 1C. A comparison between Nf for the

DCSE limit and Nf for the FE limit for all three temperatures with a 9000 lb. load is

shown in Figures 5.22 through 5.24. For all cases, Nf was lower for DSCE than for FE.

This indicates that DSCE is more critical for all of these sections.

The results of the propagation phase were also analyzed for these sections

individually and appear below in Figures 5.25 through 5.27. The Nf for the propagation

phase is defined as the number of cycles that are required to propagate the crack from an

initial length of 4mm to a final length of 50mm.











S7000 *9000 0 11000


50000

a 40000
0
S30000

S20000

10000

0 -


I75-1U I75-1C


I75-2U I75-3C
Section


SR80-2U SR80-1C


Figure 5.16 Number of Cycles to Failure for DCSE at 00 C


S7000 M 9000 ] 11000


50000

S40000

S30000
S20000

10000

0


I75-1U I75-1C I75-2U I75-3C
Section


SR80-2U SR80-1C


Figure 5.17 Number of Cycles to Failure for DCSE at 100 C


M7000 9000 0 11000


50000

40000
-4-
S30000
20000
S 20000
-*

10000

0


I1M-N-,


175-1U 175-1C


I75-2U I75-3C
Section


SR80-2U SR80-1C


Figure 5.18 Number of Cycles to Failure for DCSE at 200 C


la~a~a_








50




*7000 9000 0 11000


50000

a 40000
-

* 30000
-
S 20000
1 -
10000

0


175-1U 175-1C


I75-2U I75-3C

Section


SR80-2U SR80-1C


Figure 5.19. Number of Cycles to Failure for FE at 00 C


*7000 *9000 0 11000


50000

a 40000

A 30000

S20000

10000

0


I75-1U I75-1C I75-2U I75-3C

Section


Figure 5.20. Number of Cycles to Failure for FE at 100 C


*7000 9000 0 11000


50000

a 40000
-

* 30000
20000
10000
10000


SR80-2U SR80-1C


I75-1U I75-1C


I75-2U I75-3C

Section


SR80-2U SR80-1C


Figure 5.21. Number of Cycles to Failure for FE at 200 C








51



*DCSE M FE


I75-1U I75-1C


I75-2U I75-3C
Section


SR80-2U SR80-1C


Figure 5.22. Comparison between Nf ofDCSE and FE limits at 00 C


*DCSE FE


50000

a 40000

9 30000

S20000

10000

0


I75-1U I75-1C I75-2U I75-3C
Section


SR80-2U SR80-1C


Figure 5.23. Comparison between Nf ofDCSE and FE limits at 100 C


DCSE 0 FE


50000

a 40000

* 30000

S20000

10000

0


I75-1C I75-2U I75-3C
Section


SR80-2U SR80-1C


Figure 5.24. Comparison between Nf ofDCSE and FE limits at 100 C


50000

a 40000

* 30000

S 20000

10000 -

0-


- -


I75-1U








52




*7000 9000 0 11000


40000
35000
30000
25000
| 20000
15000
10000
5000
0


175-1U 175-1C 175-2U 175-3C


SR80-2U SR80-1C


Section


Figure 5.25. Number of Cycles to Failure to 50mm at 00 C


S7000 9000 0 11000


40000
35000
30000
25000
20000
15000
10000
5000
0


I75-1U I75-1C I75-2U I75-3C SR80-2U SR80-1C


Section


Figure 5.26. Number of Cycles to Failure to 50mm at 100 C


E7000 m9000 11000


40000
35000
o 30000
S 25000
-
S20000
15000
-
S10000
Z 5000
0


I75-1U I75-1C


I75-2U I75-3C SR80-2U SR80-1C

Section


Figure 5.27. Number of Cycles to Failure to 50mm at 200 C









For each case at all temperatures and loadings, the uncracked sections required

more cycles to propagate a crack to 50mm than their paired cracked sections. There were

significant differences in Nf between the paired sections. 1-75 section 1U required

approximately twice the cycles of section 1C to propagate to 50mm. 1-75 section 2U

required close to 2.5 times the cycles of section 3C. SR-80 section 2U required over six

times the cycles of section 1C. The values of Nf for SR-80 section 1C are the same

across all three temperatures due to the extremely high stress that this section

experienced. This caused the material to fail immediately according to the cracking

model.

5.5 Individual Analysis of the Sections

5.5.1 1-75 1U and 1C

Low dissipated creep strain energy was the primary reason for the failure of section

1-75 1C. 1C and 1U had similar structures, m-value and D1. The DCSE value of 1U was

almost two times that of section 1C. From the extraction-recovery results it was found

that 1U had a much lower air void content, slightly lower asphalt content but higher film

thickness than section 1C. The air void content of section 1C was 68% higher than

section 1U. From the gradation curves, there was no significant difference between the

two sections. Both were very similar and followed the same trend. The binder viscosity

test showed that 1C had 33% higher viscosity than 1U.

The mixture property tests showed that the MR and the creep compliance values

were close to equal. Section 1C had almost 15% higher tensile strength than 1U and the

failure strain for section 1U was almost twice that of section 1C.









5.5.2 1-75 2U and 3C

The probable causes for failure in section 3C are related to the structural effects and

low DCSE values. The stresses calculated at the bottom of the asphalt for 3C are 60%

higher than those for 2U. These higher stress values resulted from lower modulus base

and sub base layers. The extraction-recovery results show significant differences. 3C

had slightly higher AC contents and 5% greater film thickness. The gradation results

showed that while both mixtures tended to be gap graded, section 3C was slightly more

gap graded. The binder viscosity for section 3C was approximately 25% greater than for

section 2U.

The mixture test results also displayed important differences between the two

sections. While the MR values were similar, the creep compliance results at 100 sec were

slightly lower for 3C at 100 C. The tensile strength of 3C was 25% lower than that of 2U.

Also, failure strain for 2U was 45% higher than for 3C.

5.2.3 SR-80 2U and 1C

The cause for failure in section 1C was due to structural effects, low DCSE values,

and high m-values. The calculated stresses for section 1C were extremely high and were

50% higher than those for section 2U. This was at least partially due to the low thickness

of the AC layer.

The trend in air void content for these two sections is the opposite of the 1-75

sections. Section 2U has as much as 70% higher air voids than section 1C. The results of

the extraction-recovery process also show many differences. Section 2U had slightly

higher AC content and film thickness than 1C. The viscosity of the binder was 86%

greater for the uncracked section. Although the gradation plots showed a similar trend,

2U was a much finer mix than section 1C.






55


The mixture test results also showed significant differences. The MR results were

close to equal for the two sections. The creep compliance however, for section 1C was

90% greater than that of section 2U. The ultimate tensile strength of 2U was 20% higher

than 1C and the failure strain was slightly higher. The m-value and DCSE values were

also significant. For all temperatures, the cracked section had much higher m-values and

section 2U had almost 40% greater DCSE values.














CHAPTER 6
FURTHER ANALYSIS

This chapter presents the results of an analysis of the pavement sections previously

discussed as well as pavement sections studied by Garcia (2002) and Sedwick (1998).

This chapter also introduces two parameters: the minimum mixture fracture toughness as

well as the mixture fracture toughness ratio.

6.1 Section Data and Mixture Test Results

A summary of the pavement section data for the sections studied by Garcia and

Sedwick are included in Appendix F. These include traffic and the stresses calculated

from BISAR. The traffic is displayed in units of millions of ESALs/Year. The stresses

were calculated as the maximum tensile stress at the bottom of the asphalt layer. A

thorough presentation of the additional section data including layer thickness and moduli

appears in Garcia (2002) and Sedwick (1998).

A summary of the mixture test results for all sections also appears in Appendix F.

As mentioned earlier, the values for m, Do, and D1 are calculated from fitting the

following equation to the creep compliance test results:

D(t) = Do +Dt

where, D(t) is compliance at time t, Do and D1 are model parameters, and m is the m-

value. It should be noted that the D1 and m-value parameters were calculated by using a

constant Do value of 3.33x10-7 psi for all sections. This was done to provide consistent

values of D1 and m.










6.2 Mixture Fracture Toughness

The number of cycles to propagate a crack 50 mm was calculated with the HMA

crack growth model for each section. Figure 6.1 shows the comparisons of Nf. Each

uncracked section clearly requires more cycles to achieve a crack length of 50mm than its

paired cracked section. It should be noted that SR-16 sections 6U and 4C were both

pavement sections that exhibited cracking in the field. However, section 4C displayed a

greater amount of cracking. All of the cracked sections also had Nf values of less than

6000. This value was chosen as the critical Nf value that separates the cracked sections

from the uncracked.

Using the value of 6000 as the critical Nf value, relationships were developed

between DCSE and Dimax for different m-values. These are shown in Figure 6.2 for

constant values of stress and tensile strength. Dimax is defined as the D1 value that

produces an Nf value of 6000 for a given DCSEf and m-value. A minimum DCSEf value

of 0.75 KJ/m3 was used for the analysis because all sections with a lower value

performed poorly.


4.00E-06
3.50E-06
3.00E-06 m-value
S2.50E-06
0.4
2.00E-06
: t0.45
1.50E-06
t0.5
1.00E-06 ^0.55
5.00E-07 E i-a- _._t --- 0.6
0.00E+00
0 0.5 1 1.5 2 2.5
DCSE (KJ/m3)

Figure 6.1. Dimax for values of DCSE and m-value















14000


12000


10000
E
E
8000

0)

& 6000
.4-
o
z
4000


2000


0


175-1U 175-1C 175-2U 175-3C SR80- SR80- SR 16- SR 16- SR 375- SR 375- TPK 1U TPK 2C NW 39- NW 39-
2U 1C 6U 4C 1U 2C 2U 1C
Section


Figure 6.1. Nf to propagate 50mm









It was noted that the relationship presented in Figure 6.2 could be expressed using a

single function of the following form:

Dlmax) = ax DCSE x mb (6.1)

where D1(max) (1/psi) is the maximum acceptable D1 value that will achieve in good

cracking performance, DCSE is the dissipated creep strain energy limit (KJ/m3), m is the

m-value, and a and b are regression constants. The coefficient b was determined to be

2.98, while a was determined to be a function of tensile strength and tensile stress as

follows:

a = 2.99 x 10 2 -10(6.36 S)+ 2.46 x 108 (6.2)

where St (MPa) is the ultimate tensile strength and c (psi) is the tensile stress in the

asphalt layer. The HMA mixture fracture toughness (KHMA) was defined as the inverse of

D1. Therefore, a minimum fracture toughness can be defined as the inverse of D1(max)

which can be expressed as follows:

298
KHMA(n) = [2.99 x1-2 C31(6.36- )+2.46 m8 (6.3)
2.99x10 310 (6.36 S)+ 2.46 x 10- DCSE











1.4E-07

1.2E-07

1.OE-07
a (usi)
8.0E-08 -
A 100
6.0E-08 A 10

4.0E-08 150

2.0E-08

0.OE+00
0 1 2 3 4 5 6 7
St (MPa)

Figure 6.3. Relationship between a, St, and c

Furthermore, a KHMA Ratio is defined as the ratio of the mixture fracture

toughness to the minimum mixture fracture toughness as follows:

K
KHMARatio = HMA (6.4)
KHMA(mln)

where, KHMA equals 1/D1 and KHMA(min) is determined by Equation 6.3.

The KHMA Ratio allows the comparison of the cracking resistance of different

pavement sections. A mixture with a KHMA Ratio value greater than 1 will have good

cracking performance, while a mixture with a KHMA ratio value of less than 1 will have

poor cracking performance. Figure 6.4 shows a comparison of the field sections and their

respective KHMA ratios. From the figure, it can be seen that all cracked sections exhibited

a fracture toughness ratio of less than one except for those sections with a DCSEf value of

less than 0.75KJ/m3. Each of the uncracked sections exhibited a KHMA Ratio of greater

than one excluding section 1-10 MWlwhich had an unusually large DCSEf value.






61


Figure 6.5 displays the KHMA ratios for all sections with DCSEf values between

0.75 KJ/m3 and 2.5KJ/m3. For this range of DCSEf values, it appears that the KHMA ratio

is accurate in predicting the cracking performance of the pavement sections.


















L.uu

1.50

1.00

0.50

0.00

9 4\p3 cl &02

nSS ^^^^^ 6' ^L~ ^\ ^ ^^ /'


Cracked


Uncracked


Section


note: DCSE<0.75 KJ/m3 + DCSE>2.5KJ/m3


Figure 6.3. KHMA Ratio for all sections















--I I I I


-_=.... 111


111111


US SR80- 110- 110- SR 16- 110- SR 16- NW TPK I75-3C I75-1C SR I75-1U I75-2U TPK SR80- NW SR 195-
301- 1C DE DW 4C MW2 6C 39-1C 2C 375- 1U 2U 39-2U 375- DN


Section


Uncracked


Figure 6.4. KHMA Ratio for sections 0.75 KJ/m3 < DCSEf < 2.5 KJ/m3


2.50
.2 200
< 1.50
1.00
0.50
0.00


Cracked









6.3 Mixture Properties

An analysis was performed on all of the pavement sections to identify the key

mixture properties that affect cracking performance. Appendix F displays the various

mixture properties for each section including binder viscosity, effective asphalt content,

theoretical film thickness, percent air voids, and VMA. From these figures, it is clear that

no clear relationship exists between any of these properties and the relative cracking

performance.

However, several trends were observed when analyzing the gradations of each

section. The KHMA ratios were calculated for each section with a constant stress of 120

psi. This was done to eliminate the effect of structure in the calculation. The gradations

were then grouped according to high and low KHMA ratios as well as mix designation to

observe any trends between gradation characteristics and KHMA ratio. These gradations

are shown for all sections in Figures 6.5 through 6.8 below.

Two primary differences in gradation are apparent between the low KHMA ratio

sections and the high KHMA sections. The low KHMA sections move away and remain

further away from the max density line. These sections either remain fine relative to the

line or approach it and then gap dramatically at the finer sieves. The gradation curves for

the high KHMA sections are parallel and remain close to the max density line. They also

appear coarser with respect to the line without gapping drastically at the finer sieves. A

summary of these trends is illustrated in Figures 6.9-6.12.



















100


90


80


70




50
" 50 0.25 SR16-4C

0.42
40 -- SR16-6C
0.38

0.30

20 --- -I-10 DW

Max Density
10 Line



0 200 100 50 30 16 8 4 3/8 1/2 3/4 1
Sieve sizes


Figure 6.5. Gradations of Low KHMA Ratio Sections with 12.5mm Nominal Aggregate Size



















100


90


80


70


60


S50-
2.10 SR375-1U
40 3.12 -0- SR80-2U

0.86 -~-I-95 DN
30 / 0.88 -- SR3752C
1.21 ---NW391C
20 0.90 -E-I75 3C


10 Max Density
Line

0
0 200 100 50 30 16 8 4 3/8 1/2 3/4 1
Sieve sizes


Figure 6.6. Gradations of High KHMA Ratio Sections with 12.5mm Nominal Aggregate Size




































U KH Ratio
90


80


70


60


50


40 0.05 --US301 BS
0.05
0.69 ----US301 BN
30 0.22 ----I-10 DE
0.48 -I-1-10 MW2
20
S Max Density
Line
10


0
0 200 100 50 30 16 8 4 3/8 1/2 3/4 1
Sieve sizes


Figure 6.7. Gradations of Low KHMA Ratio Sections with 9.5mm Nominal Aggregate Size


















100

90


80


70 Max Density
Line
60 KHA Ratio

1.40
A 50
3.89 XTPKU
NW 39-2U
40 0 0 0 .9 7
1.78 --I75-2U
30 1.26 --175-1U







0
0.91 -0-1-10 MW1
20 0.97 -W-TPK2C
10 ---- 175 IC




0 200 100 50 30 16 8 4 3/8 1/2 3/4 1
Sieve sizes


Figure 6.8. Gradations of High KHMA Ratio Sections with 9.5mm Nominal Aggregate Size





































200 100 50 30


16 8


4
Sieve sizes


3/8 1/2


Figure 6.9. Case 1 Gradation Comparison for 12.5mm Nominal mixes


0 200 100 50 30 16 8 4 3/8 1/2 3/4
Sieve sizes



Figure 6.10. Case 2 Gradation Comparison for 12.5mm Nominal mixes


'-<








"-- LowKHMA -



High KHMA -

Max Density -
m- Line








70





100

90









P 40- -" = Low KHMA

30
S- High KHMA

1 Max Density
Line

0 200 100 50 30 16 8 4 3/8 1/2 3/4

Sieve sizes

Figure 6.11. Case 1 Gradation Comparison for 9.5mm Nominal mixes




100

90

80 -







S40 High KHMA

30
*- -* Low KHMA
20
10 -- Max Density
Lme


0 200 100 50 30 16 8 4 3/8 1/2 3/4
Sieve sizes


Figure 6.12. Case 2 Gradation Comparison for 9.5mm Nominal mixes









6.4 Traffic

Traffic has an important effect on the initiation and propagation of surface cracks in

pavements. A higher level of traffic generally means a higher number of load repetitions

and a higher likelihood of high critical loads. It may be important to require higher

minimum KHMA ratios for higher traffic sections during the design process.

It was assumed that a road with a higher volume of traffic would require a

pavement system that would provide a higher number of cycles until a crack develops.

Average values of c, St, DCSEf, m-value, and D1 were chosen that would result in 6000

Nf at a crack length of 2 in. or a KHMA Ratio of 1. The D1 value was varied while all

other values were held constant to produce different KHMA ratios. The resulting Nf values

were calculated using the HMA fracture mechanics model. Figure 6.13 below shows the

relationship between KHMA Ratio and Nf. Nf is directly proportional to traffic loading.

Figure 6.13 shows that an increase in traffic by a factor of two will require a pavement

system with two times the KHMA ratio.


2.5

2

I 1.5



0.5


0 2000 4000 6000 8000 10000 12000 14000
Nf

Figure 6.13. Relationship Between Nf and Required KHMA Ratio









From the relationship between Nf and KHMA ratio, the factor FT was defined. This

factor accounts for the increased KHMA ratio that would be required for increased traffic.

This relationship between traffic level and required KHMA ratio was compared to actual

field section data. The uncracked section with the lowest KHMA ratio in this study was

section I75-1U with a value of 1.27. This section also possessed a traffic loading of

approximately 500,000 ESALS/year. The relationship shown in Figure 6.13 above was

calibrated to this value where at a traffic level of 500,000 ESALS/year, the FT factor

equals 1.3. Figure 6.14 below shows FT VS. traffic level.


3

2.5

2

S1.5

1

0.5


0 200 400 600 800 1000 1200
Traffic (ESALS/year xl000)

Figure 6.14. FT vs. Traffic loading (ESALS/year x 1000)

This calculation does not however consider the structural differences that typically

exist between pavement sections with considerably different traffic loading. An analysis

was performed on theoretical pavement sections with differing traffic loadings. Several

pavement sections corresponding to different traffic loads were generated using the

AASHTO pavement design method. These were generated for sections with traffic

ranging from 50,000 ESALS/year to 1x106 ESALS/year. The design life of the pavement

was assumed to be 20 years. Constant values of layer moduli, reliability, and standard










deviation were assumed. A sample calculation appears in Appendix G. The layer

thicknesses were calculated as well as the maximum tensile stress at the bottom of the

AC layer. Using these stresses, the resulting KHMA ratios were calculated. Pavement

sections with higher traffic loads resulted in thicker AC layers as well as lower stresses at

the bottom of the AC layer and therefore possessed higher KHMA ratios. The factor Fs

considers the effect of the increase in structural capacity on the increase in KHMA ratio.

The resulting relationship between traffic loading and Fs is shown in Figure 6.15 below.


1.2



0.8

S0.6

0.4

0.2

0 -
0 200 400 600 800 1000 1200
Traffic (ESALS/year xO000)

Figure 6.15. Relationship Between Traffic Level and Fs

The Fs factor was also calibrated to the field section I75-1U with a traffic level of

approximately 500,000 ESALS/year. Figure 6.16 below shows the resulting relationship.











6

5

4



2
1 -

0
0 200 400 600 800 1000 1200
Traffic (ESALS/year x 1000)

Figure 6.16. Fs vs. Traffic (ESALS/year x 1000)

Considering the effect of traffic as well as the change in structure results in the

following equation:

KHMARatio Re quired = F, Fs (6.5)

where KHMA Ratio Required is the minimum required KHMA ratio, FT is the traffic

factor, Fs is the structural factor, and KHMA Ratio is the ratio of actual KHMA to the

minimum required KHMA. Figure 6.17 shows the minimum required KHMA ratio for

different traffic loads. The value of 1 was used for sections with traffic levels less than or

equal to 250k ESALS/yr.












2.5

2

1.5

1

S0.5

0
0 200 400 600 800 1000 1200
Traffic (ESALS/year x1000)


Figure 6.17 Minimum KHMA Ratio Required vs. Traffic Levels














CHAPTER 7
SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

7.1 Summary and Conclusions

The findings of this study may be summarized as follows:

* The important effects of mixture properties and pavement structure on top-down
cracking performance was verified.

* It appears that HMA fracture mechanics properly accounts for effects of mixture
properties. The relative cracking performance predicted by thee HMA fracture
model developed at the University of Florida agrees with field observations

* In some cases excessively low DCSEf appeared to control cracking. It appears that
cracking develops within relatively few cycles in these cases, such that cumulative
DCSEf may not appropriately represent be the mode of failure.

* A single parameter (KHMA (min)) was identified and defined that allows designers to
determine whether a mixture will experience top-down cracking. This parameter
accounts for the effects of both mixture properties and pavement structure. It is
based on MR, Creep, and Strength test results.

* The following relationship was developed for KHMA(min) based on calibration to
actual field cracking performance of mixtures:

K 298
KHM(m) [2.99x 10 2 310(6.36- S)+ 2.46x 10 ]DCSE

* Furthermore, a mixture fracture toughness ratio was defined that allows relative
comparisons between different mixtures and pavement sections:

K 1
KHMRatio = HA where, KH and
KHMA(mn) D1

If KRatio > 1 No Cracking

If KRatio < 1 Cracking

* The KHMA Ratio was shown to predict field cracking performance for all field
sections evaluated (22) except for three sections with very low or very high DCSEf.










* Specific gradation characteristics were associated with poor cracking performance.
It was concluded that the KHMA Ratio may be increased by changes in the gradation
such as the degree to which the gradation curve is parallel to the max density line
as well as the severity of gap grading in the finer sieves.

* A procedure was established to rationally account for the effects of traffic on the
minimum required KHMA Ratio. The following recommendations resulted from the
work:

Traffic Minimum
ESALS/year x 1000 KHMA Ratio
<250 1
400 1.2
500 1.3
1000 1.95

7.2 Recommendations

The specific condition and mechanism associated with top-down cracking is yet to

be determined. The relations developed were based on relative tensile stress as

determined at the bottom of the asphalt concrete layer. The critical tensile stresses at the

top of the pavement may involve the introduction of a crack and a temperature gradient

into the calculation as well as the residual stresses that may be induced by creep.















APPENDIX A
SUMMARY OF NON-DESTRUCTIVE TESTING (FWD)









Table A. 1: Deflection (milsO From 1-75 1U
Sensor Spacing (in)
Milepost 0 8 12 18 24


2.01 5.53 4.42 3.72 2.91 2.28 1.44 0.8 9142
2.02 5.95 4.89 4.19 3.22 2.49 1.5 0.76 9002
2.03 5.8 4.71 4 3.09 2.38 1.45 0.76 9086
2.04 5.98 4.82 4.11 3.17 2.43 1.54 0.85 9094
2.05 5.33 4.38 3.79 2.93 2.32 1.59 0.87 9121
2.06 5.46 4.35 3.69 2.89 2.3 1.46 0.85 8954
2.07 5.78 4.66 3.91 3.15 2.43 1.51 0.87 9089
2.08 6.09 4.91 4.2 3.25 2.5 1.6 0.89 9050
2.09 6.04 4.89 4.19 3.26 2.448 1.53 0.85 9007
2.1 5.63 4.58 3.89 3 2.36 1.52 0.86 8938


Table A.2: Deflection milss) From 1-75 1C
Sensor Spacing (in) Load
Milepost 0 8 12 18 24 36 60 Lbs.
1.01 6.41 5.15 4.32 3.33 2.57 1.63 1.01 8943
1.02 6.56 5.19 4.34 3.26 2.55 1.6 0.91 8803
1.03 6.28 5.01 4.21 3.15 2.46 1.62 0.94 8911
1.04 6.27 5.07 4.29 3.26 2.56 1.66 1 8891
1.05 6.38 5.17 4.37 3.36 2.65 1.69 0.96 8856
1.06 6.39 5.22 4.47 3.51 2.74 1.77 1 8771
1.07 6.3 5.16 4.4 3.47 2.77 1.79 0.95 8681
1.08 6.38 5.23 4.43 3.5 2.79 1.81 1.05 8811
1.09 6.15 5.05 4.32 3.43 2.71 1.82 1.01 8800
1.1 6.87 5.64 4.79 3.76 2.94 1.85 1 8800


Load
Lbs.


















Sensor Spacing (in)
0 10 20 30 40 50 60 70
0


2


4


--6 ---2.01
--i- 2.02
8 2.03
2.04
"- -2.05
S10
--- 2.06
2.07
12 2.08

2.09
14 2.1


16


18


20


Figure A. 1 Deflections for 1-75 1U


















Sensor Spacing (in)
0 10 20 30 40 50 60 70
0


2


4


6 --- 1.01
--i- 1.02
8 1.03
1.04
) -- 1.05
S10
S-0- 1.06

-5--- 1.07
12 1.08

1.09
14 1.1


16


18


20


Figure A.2 Deflections for 1-75 1C









Table A.3: Deflections for 1-75 2U


Milepost


Sensor Spacing (in)


Load
lbs.


1.01 4.69 3.74 3.24 2.68 2.31 1.65 0.99 8883
1.02 4.39 3.5 3.03 2.5 2.08 1.5 0.91 9046
1.03 4.67 3.52 3.02 2.5 2.08 1.49 0.91 9094
1.04 4.4 3.45 2.95 2.41 2.07 1.57 0.92 8792
1.05 4.49 3.37 2.88 2.41 1.99 1.49 0.9 9022
1.06 4.73 3.81 3.28 2.72 2.25 1.66 0.91 9007
1.07 4.83 3.87 3.26 2.72 2.26 1.62 0.94 9026
1.08 4.78 3.79 3.17 2.61 2.14 1.47 0.81 8954
1.09 4.68 3.83 3.31 2.67 2.26 1.74 1.02 9054
1.1 4.64 3.68 3.09 2.57 2.18 1.61 0.93 8962


Table A.4: Deflections for 1-75 3C
Sensor Spacing (in) Load
Milepost 0 8 12 18 24 36 60 lbs.
2.01 6.16 5.01 4.13 3.24 2.5 1.52 0.79 8943
2.02 6.16 4.85 4.11 3.2 2.52 1.62 0.8 8962
2.03 6.34 4.91 4.14 3.23 2.54 1.63 0.83 8943
2.04 5.89 4.83 4.02 3.21 2.43 1.57 0.8 8819
2.05 6.5 5.08 4.15 3.22 2.44 1.52 0.8 8927
2.06 6.69 5.41 4.58 3.59 2.8 1.75 0.82 8840
2.07 6.59 5.41 4.52 3.47 2.68 1.66 0.82 8859
2.08 6.36 5.04 4.24 3.31 2.56 1.63 0.8 8856
2.09 6.24 4.99 4.12 3.25 2.51 1.57 0.82 8803
2.1 5.94 4.72 3.93 3.07 2.44 1.56 0.82 8851






















Sensor Spacing (in)
0 10 20 30 40 50 60 70
0


2


4


6

-- 1.01
08 ---1 1.02
1.03
1.04 o0
10
U-*- 1.05
E-- 1.06
12 i 1.07
-1.08
14 1.09
1.1

16


18


20


Figure A.3 Deflections for 1-75 2U




















Sensor Spacing (in)


Figure A.4 Deflections for 1-75 3C


4


6


S8


0 10


S12


14


16


18


20


* 2.01
---- 2.02
2.03
2.04
W- 2.05
--- 2.06
-I- 2.07
--2.08
2.09
2.1


~L1







L









Table A.5: Deflections for SR-80 2U


Milepost


Sensor Spacing (in)


Load
lbs.


2.01 7.48 5.77 4.97 4.1 3.39 2.18 1 9054
2.02 7.15 5.98 5.22 4.27 3.5 2.27 1.07 8906
2.03 7.59 6.48 5.55 4.46 3.66 2.42 1.14 8800
2.04 8.55 6.7 5.53 4.41 3.59 2.35 1.13 8967
2.05 9.79 7 5.94 4.6 3.62 2.19 1.07 8864
2.06 7.25 6.18 5.36 4.37 3.52 2.24 1.14 9018
2.07 7.68 6.2 5.35 4.32 3.45 2.19 1.07 8994
2.08 8.16 6.73 5.81 4.58 3.73 2.39 1.21 8970
2.09 8.51 6.26 5.33 4.07 3.23 2.02 1.09 8720
2.1 9.35 6.71 5.64 4.38 3.5 2.16 1.11 8668


Table A.6: Deflections for SR-80 1C
Sensor Spacing (in) Load
Milepost 0 8 12 18 24 36 60 lbs.
1.01 10.99 7.75 5.44 3.33 2.26 1.21 0.76 8613
1.02 10.91 7.49 5.44 3.41 2.24 1.32 0.81 8450
1.03 10.31 7.02 5.07 3.02 2.07 1.19 0.78 8414
1.04 9.88 6.78 4.85 3.11 2.09 1.31 0.83 8593
1.05 11.75 8.24 5.78 3.52 2.27 1.26 0.83 8477
1.06 14.06 10.22 7.61 5.07 3.51 2.17 1.29 8315
1.07 17.67 12.91 9.74 6.54 4.56 2.72 1.51 8307
1.08 18.84 13.7 10.13 6.56 4.63 2.85 1.59 8347
1.09 17.02 12.54 9.62 6.68 4.85 2.91 1.63 8334
1.1 17.12 13.12 10.34 7.44 5.37 3.22 1.71 8318


















Sensor Spacing (in)
20 30


Figure A.5 Deflections for SR-80 2U


4


6


- 8


0 10


0 12


14


16


18


20


-- 2.01
-I-- 2.02
2.03
2.04
W- 2.05
--- 2.06
-I- 2.07
--2.08
2.09
2.1