<%BANNER%>

Evaluation of Open-Graded and Bonded Friction Course for Florida


PAGE 1

EVALUATION OF OPEN-GRADED AND BONDED FRICTION COURSE FOR FLORIDA By ARVIND VARADHAN 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 2004

PAGE 2

ACKNOWLEDGMENTS I would like to thank Dr. Bjorn Birgisson and Dr. Reynaldo Roque for their guidance throughout the project. I believe that their knowledge and expertise helped me understand and solve crucial problems during the course of the project. I really appreciate the technical support and advice I received from Georg Lopp throughout my research work. I would like to thank Lokendra, Claude, Tanya and Eddy for their assistance in performing various laboratory tests. I would also like to thank Christos Drakos for helping me with KENLAYER in performing stress analysis. I would like to thank Greg Sholar, Shanna Jhonson and Ricky Lloyd from the research wing of DOT for their help during the course of the project. I would like to thank all my friends for providing an unforgettable and enjoyable time during my two years of study in Gainesville. Finally, I would like to thank my parents and my brother for all the love and support they given me throughout my academic years. ii

PAGE 3

TABLE OF CONTENTS Page ACKNOWLEDGMENTS..................................................................................................ii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.........................................................................................................viii ABSTRACT......................................................................................................................xii CHAPTER 1 INTRODUCTION .........................................................................................................1 1.1 Thick Open Graded Friction Courses.....................................................................1 1.2 Bonded Friction Course Evaluation........................................................................2 1.3 Objective.................................................................................................................3 1.3.1 FC-5 and Novachip.......................................................................................3 1.3.2 Georgia PEM................................................................................................4 2 LITERATURE REVIEW ..............................................................................................5 2.1 Novachip.................................................................................................................5 2.1.1 Novachip Description...................................................................................5 2.1.2 Paving Equipment........................................................................................6 2.1.3 Post Construction Testing.............................................................................7 2.1.3.1 Surface roughness International roughness index...........................7 2.1.3.2 Skid friction........................................................................................7 2.1.3.3 Surface macrostructure.......................................................................8 2.1.3.4 Ride quality data.................................................................................8 2.1.3.5 Rolling Noise......................................................................................9 2.1.4 NOVACHIP Mix Design.............................................................................9 2.1.4.1 Gradation............................................................................................9 2.1.4.2 Asphalt content determination...........................................................9 2.1.5.1 Surface treatment paving machine...................................................11 2.1.5.2 Rollers............................................................................................13 2.1.5.3 Straightedges and templates.............................................................13 2.1.5.4 Weather limitations..........................................................................13 2.1.5.5 Tack coat..........................................................................................13 2.1.5.6 Hauling equipment...........................................................................14 2.1.5.7 Spreading and finishing....................................................................14 iii

PAGE 4

2.1.5.7 Compaction......................................................................................14 2.1.5.8 Method of measurement...................................................................15 2.2 Porous European Mix (PEM)...............................................................................15 2.2.1 Advantages of Porous Asphalt...................................................................15 2.2.1.1 Hydroplaning and glare reduction....................................................15 2.2.1.2 Noise reduction................................................................................16 2.2.1.3 Skid friction......................................................................................17 2.2.2 Disadvantages of Porous Asphalt...............................................................17 2.2.2.1 Strength............................................................................................17 2.2.2.2 Initial stiffness modulus...................................................................18 2.2.2.3 Aging and stripping..........................................................................18 2.2.2.4 Temperature.....................................................................................18 2.2.2.5 Clogging...........................................................................................19 2.2.3 Performance Related Laboratory Testing...................................................19 2.2.3.1 Resistance to plastic deformation.....................................................19 2.2.3.2 Resistance to indirect tension...........................................................20 2.2.3.3 Resistance to disintegration..............................................................20 2.2.3.4 Adhesiveness....................................................................................20 2.2.3.5 Drainage test.....................................................................................21 2.2.4 Mix Design Approach................................................................................21 2.2.4.1 British design....................................................................................22 2.2.4.2 Spanish design..................................................................................22 2.2.4.3 Italian design....................................................................................23 2.2.4.4 Belgium design.................................................................................23 2.3 Georgia PEM........................................................................................................24 2.3.1 Material Selection.......................................................................................23 2.3.1.1 Composition.....................................................................................23 2.3.1.2 Gradation..........................................................................................26 2.3.1.3 Aggregate specification....................................................................27 2.3.1.4 Polymer modified asphalt cement....................................................27 2.3.1.5 Mineral fibers...................................................................................28 2.3.2 Georgia PEM mix design procedure ..................................................28 3 MATERIALS..............................................................................................................31 3.1 FC-5 and Novachip.............................................................................................31 3.1.1 Aggregates................................................................................................31 3.1.2 Binder .......................................................................................................31 3.2 Georgia PEM ......................................................................................................32 3.2.1 Aggregate and Binder ..............................................................................32 4 DETERMINATION OF COMPACTION LEVEL AND AGING PROCEDURE FOR FRICTION COURSES...............................................................................................34 4.1 Compaction of the Friction Course Mix.............................................................34 4.1.1 Compaction Data......................................................................................34 4.1.2 Initial Study ..............................................................................................38 iv

PAGE 5

4.2 Asphalt Content Determination..........................................................................44 4.3 Mixing and Compaction .....................................................................................45 4.4 Long-Term Oven Aging of Friction Course .......................................................47 5 INDIRECT TENSILE TESTING................................................................................49 5.1 Sample Preparation .............................................................................................50 5.2 Test Procedures...................................................................................................53 5.2.1 Resilient Modulus Test.............................................................................53 5.2.2 Creep Test.................................................................................................55 5.2.3 Strength Test............................................................................................56 5.3 Issues Related with IDT Testing of Friction Course ..........................................59 6 MIX DESIGN APPROACH FOR GEORGIA PEM...................................................61 6.1 Evaluation of Compaction Level for Georgia PEM ...........................................61 6.2 Mix Design Approach.........................................................................................61 7 FINDINGS AND ANALYSIS ....................................................................................65 7.1 Evaluation of FC-5 and Novachip Field Mix .....................................................66 7.1.1 Unaged Field Mix.....................................................................................67 7.1.2 Aged Field Mix.........................................................................................67 7.1.3 Saturated Unaged Mix..............................................................................74 7.1.4 Saturated Aged Mix..................................................................................75 7.1.5 Moisture Conditioning..............................................................................83 7.2 Evaluation of Georgia PEM................................................................................88 8 SUMMARY, FINDINGS AND ANALYSIS..............................................................94 8.1 Summary of Findings .........................................................................................94 8.2 Conclusions.........................................................................................................95 8.3 Recommendations...............................................................................................96 APPENDIX A VOLUMETRICS FOR FC-5 AND NOVACHIP.......................................................97 B VOLUMETRICS FOR GEORGIA PEM .................................................................101 C IDT TEST RESULTS FOR FC-5, NOVACHIP AND GEOGIA PEM ...................107 LIST OF REFERENCES.................................................................................................116 BIOGRAPHICAL SKETCH...........................................................................................118 v

PAGE 6

LIST OF TABLES Table Page 2.1 International Roughness Index......................................................................................7 2.2 Skid Number..................................................................................................................8 2.3 Ride Quality...................................................................................................................9 2.4 JMF Gradation Range..................................................................................................10 2.5 Georgia PEM Gradation..............................................................................................26 2.6 Aggregate Specifications for Georgia PEM................................................................27 2.7 Design Requirements for Georgia PEM......................................................................28 3.2 Mixes from the Field....................................................................................................32 4.1 Air Voids......................................................................................................................35 4.2 Locking point For the Mixtures...................................................................................38 4.3 Locking Point Based on Gradient of Slope..................................................................41 4.4 Locking Points of all Mixtures based on Gradient of Slope........................................41 4.5 Air Voids for 50 Gyrations..........................................................................................42 4.7 Weight of PG 76-22 to be added..................................................................................47 6.1 Locking Point for Georgia PEM..................................................................................62 A.1 Gradation for FC-5 Granite, Limestone and Novachip...............................................98 A.2 Rice Test for the Friction Courses..............................................................................99 A.3 Air Voids and Asphalt Content for Friction course..................................................100 B.1 Gradation for Georgia PEM......................................................................................102 B.2 Bulk Specific Gravity for Georgia PEM...................................................................103 vi

PAGE 7

B.3 Rice Test for Georgia PEM.......................................................................................103 B.4 Drain-down Test for Georgia PEM...........................................................................104 B.5 Film Thickness for Georgia PEM.............................................................................105 C.2 Summary of Fracture Results for Saturated Mix......................................................109 C.3 Summary of Fracture Results for Field Mix and Georgia PEM...............................110 C.4 Summary of Fracture Results for Different Aged Fc-5 Lime Mixes........................111 vii

PAGE 8

LIST OF FIGURES Figure page 2.1 Surface Constant Kc vs. Percentage Oil Retained.......................................................12 2.2 British Gradations .......................................................................................................24 2.3 Spanish Gradation........................................................................................................25 2.4 Italian Gradation Band.................................................................................................26 2.5 Georgia PEM Gradation Band.....................................................................................27 3.1 Gradation of FC-5 Granite, FC-5 Limestone, Novachip and Georgia PEM...............33 4.1 Compaction Curve for FC-5 Limestone......................................................................36 4.2 Compaction Curve for NOVACHIP............................................................................36 4.3 Compaction for FC-5 Granite......................................................................................37 4.4 Locking Point for FC-5 with Limestone by Visual Observation.................................39 4.5 Locking Point for Novachip by Visual Observation....................................................39 4.6 Locking Point for FC-5 with Granite by Visual Observation......................................40 4.7 Gradations after Extraction .........................................................................................43 4.8 1/8 Mesh used for Containing the Pill.......................................................................48 4.9 Pill Contained with Mesh.............................................................................................48 5.1 Effect of Loading Condition........................................................................................49 5.2 Cutting Device.............................................................................................................51 5.3 Gauge Point Attachment..............................................................................................52 5.4 Marking Loading Axes................................................................................................53 5.5 Power Model for Creep Compliance...........................................................................57 viii

PAGE 9

5.6 FE and DCSE from Strength Test................................................................................59 6.1 Final Mix Design for Georgia PEM.............................................................................64 7.1 Typical Pavement Structure for Stress Calculation.....................................................66 7.2 Fracture Energy for Field Mix.....................................................................................68 7.3 Dissipated Creep Strain Energy for Field Mix............................................................68 7.4 Energy Ratios for Field Mix........................................................................................69 7.5 Failure Strain for Field Mix.........................................................................................69 7.6 Tensile Strength for Field Mix.....................................................................................70 7.7 Resilient Modulus for Field Mix..................................................................................70 7.8 Creep Rate for Field Mix.............................................................................................71 7.9 Aging Effect in Fc-5 Granite and FC-5 Limestone.....................................................73 7.10 FE for FC-5 Lime at Various Stages of Aging..........................................................75 7.11 DCSE for FC-5 Lime at Various Stages of Aging.....................................................76 7.12 Failure Strain for FC-5 Lime at Various Stages of Aging.........................................76 7.13 Energy Ratio for FC-5 Lime at Various Stages of Aging..........................................77 7.14 Tensile Strength for FC-5 Lime at Various Stages of Aging.....................................77 7.15 Resilient Modulus for FC-5 Lime at Various Stages of Aging.................................78 7.16 Creep Rate for FC-5 Lime at Various Stages of Aging.............................................78 7.17 Comparison of ER between unaged Saturated and Field Mix...................................79 7.18 FE for Saturated Mix.................................................................................................79 7.19 DCSE for Saturated Mix............................................................................................80 7.20 Failure Strain for Saturated Mix................................................................................80 7.21 ER for Saturated Mix.................................................................................................81 7.22 Tensile Strength for Saturated Mix............................................................................81 7.23 Resilient Modulus for Saturated Mix.........................................................................82 ix

PAGE 10

7.24 Creep Strain Rate for Saturated Mix..........................................................................82 7.25 Creep Deformation Curve for Conditioned samples of FC-5 with Granite...............84 7.26 Creep Rate for Unconditioned and Conditioned samples FC-5 Granite....................85 7.27 FE after Conditioning for FC-5 with Limestone........................................................85 7.28 DCSE after Conditioning for FC-5 with Limestone..................................................86 7.29 ER after Conditioning for FC-5 with Limestone.......................................................86 7.30 Failure Strain after Conditioning for FC-5 with Limestone......................................87 7.31 Mr after Conditioning for FC-5 with Limestone.......................................................87 7.32 Creep Rate after Conditioning for FC-5 with Limestone..........................................88 7.32 Loosely and Well Reinforced Matrix of Aggregate and Asphalt..............................89 7.33 FE for Georgia PEM..................................................................................................90 7.34 DCSE for Georgia PEM.............................................................................................90 7.35 Failure Strain for Georgia PEM.................................................................................91 7.36 ER for Georgia PEM..................................................................................................91 7.37 Tensile Strength for Georgia PEM............................................................................92 7.38 Resilient Modulus for Georgia PEM.........................................................................92 7.39 Creep Rate for Georgia PEM.....................................................................................93 B.1 Final Mix Design for Georgia PEM..........................................................................106 C.1 m value for Saturated Field Mix...............................................................................112 C.2 D1 value for Saturated Field Mix..............................................................................112 C.3 m value for Field Mix and Georgia PEM..................................................................113 C.4 D1 value for Field Mix and Georgia PEM................................................................113 C.5 m value for FC-5 Lime at Different Stages of Aging...............................................114 C.6 D1 value for FC-5 Lime at Different Stages of Aging..............................................114 C.7 m value for conditioned FC-5 Lime..........................................................................115 x

PAGE 11

C.8 D1 Value for Aged conditioned FC-5 Lime..............................................................115 xi

PAGE 12

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 EVALUATION OF OPEN-GRADED AND BONDED FRICTION COURSE FOR FLORIDA By Arvind Varadhan August 2004 Chair: Bjorn Birgisson Cochair: Reynaldo Roque Major Department: Civil and Coastal Engineering The project involves the fracture evaluation of 1) FC-5 with Granite and limestone (Friction course used in Florida) and NOVACHIP (bonded friction course) obtained from the field, and 2) Georgia PEM prepared in the lab. As a part of the study, an extensive literature review was done covering the various kinds of friction courses used around the world. The focus of literature review was on various mix design approaches for thick open graded and bonded friction courses and materials, equipment and construction guidelines as specified in different states in US. The field mixes obtained from the US 27 Highway Project were tested in the laboratory to evaluate their fracture properties. In addition, the effect of aging and moisture damage on their fracture resistance was also evaluated. The first step in the process was the identification of the compaction level for these friction courses. Thus, the locking point for the friction course was identified based on our study of the rate of change of compaction. xii

PAGE 13

To replicate the interface, consisting of Novabond, between the underlying HMA and the friction course, it was decided that PG 76-22 should be added to the field mix since the tack coat essentially contained PG 76-22. However, there were problems related to the workability of the mixes at such high asphalt content, which were solved by determining the compaction and mixing temperature for these mixes. Finally, the fracture testing was done using Superpave IDT test and the fracture properties were evaluated using the framework of HMA fracture mechanics. The issues concerning the IDT testing on the friction course were also identified and dealt with. In the case of Georgia PEM, the primary objective was to compare its fracture performance with FC-5 limestone and granite. The Georgia PEM mixes were prepared using the same mix design as used by Georgia DOT. Both unaged and aged mixes were evaluated for their fracture performance using IDT testing and the results were compared with FC-5 and Novachip. xiii

PAGE 14

CHAPTER 1 INTRODUCTION This project focuses on two different open graded friction course approaches: The first one is the evaluation of thick open graded friction courses for Florida conditions, and the second one is the evaluation of bonded friction courses in Florida. In the following the motivation and background for studying each of these materials has been provided and the objectives of the research project have been outlined. 1.1 Thick Open Graded Friction Courses Because of the frequency of high intensity short-duration rainfall events in Florida, vehicular hydroplaning is a serious concern, especially at the Interstate and other limited access facilities. Along with pavement cross-slope and rutting, the surface texture of an asphalt pavement plays a critical role in the prevention of hydroplaning on high-speed, multi-lane facilities. In order to minimize problems of this nature, a number of states (including Florida) have utilized traditional open-graded friction courses in order to maximize the pavements macro texture, which in turn reduces splash and spray, and minimizes hydroplaning potential. In the mid-1970, Florida developed a friction course (FC-2) which was a 3/8-inch Nominal Maximum Aggregate Size (NMAS) open-graded mixture (with polish resistant aggregate) placed approximately 1/2-inch thick. In the1990s, the FC-2 was eventually replaced by a slightly coarser open-graded friction course (FC-5), which is a 1/2 inch NMAS mixture, placed approximately inch thick. 1

PAGE 15

2 Since the FC-5 mixture is coarser and is placed slightly thicker than the FC-2, it has a greater capacity to store/drain water from the pavement surface during a severe rainstorm. However, this additional storage capacity has never been quantified. Visual observations of this surface type indicate that the while it is better than the old FC-2, it will still fill-up with water resulting in water ponding on the pavement surface. A number of European countries, as well as several states in the US (Georgia, Oregon) have developed a porous friction course which is an open-graded friction course placed in thicknesses ranging from 11/4 to 2 inches thick. The combination of the high in-place air void content of this mixture, coupled with the thickness at which it is placed, gives this type of pavement surface a great deal of potential with regard to storage and drainage of water during a rainstorm. However, there are also a number of questions associated with these types of pavements, such as fracture resistance, rutting resistance, and long-term porosity. 1.2 Bonded Friction Course Evaluation There is a high priority need to evaluate the estimated performance life and cost effectiveness of a bonded friction course that uses a special paver to lay a heavy polymer-modified tack coat just in front of the hot mix mat. This technology is available in Florida and has been demonstrated on several small projects. Other states, including Texas, Pennsylvania, and Alabama have several years experience with this process and indications are that it has great potential. The ongoing Longitudinal Wheel path Cracking study at UF has confirmed that a significant amount of the distress on high-volume Florida pavements is caused by surface initiated wheel path cracking from lateral stresses generated by radial truck tires. A recent evaluation of the ten year performance of ground tire rubber in an open graded

PAGE 16

3 friction course on SR 16 in Bradford County indicates that increased asphalt content from ground tire rubber modification of the friction course can reduce longitudinal wheel path cracking. Based on these studies, there is a high likelihood that the bonded friction course process could significantly extend the crack resistance life of open graded mixes in Florida by providing more polymer modified asphalt to the friction course. 1.3 Objective This study involves the fracture evaluation of 1) FC-5 (Friction course used in Florida) and Novachip (bonded friction course) obtained from field, and 2) Georgia PEM prepared in lab. As a part of the study, an extensive literature review was done covering the various kinds of friction courses used around the world. The focus of literature review was on various mix design approaches for thick open graded and bonded friction courses and materials, equipment and construction guidelines used by the different states. 1.3.1 FC-5 and Novachip In July 2003, five test sections were laid on US 27 highway in Highlands County of South Florida. These test sections were: 1) FC-5 Limestone, 2) FC-5 Limestone with Novabond, 3) FC-5 Granite, 4) FC-5 Granite with Novabond, and 5) NOVACHIP. The field mixes of the above test sections were tested in the laboratory to evaluate their fracture properties. In addition, the effect of moisture conditioning on the fracture resistance was also evaluated for FC-5. The first step was to identify the compaction level for these friction courses. Hence, it was important to identify the locking point for these mixes, which is defined as the point beyond which the resistance to compaction increases significantly. The locking point for the friction course was identified based on our study of the rate of change of compaction.

PAGE 17

4 To replicate the interface, consisting of Novabond, between the underlying HMA and the friction course, it was decided that PG 76-22 be added to the field mix since the tack coat essentially contained PG 76-22. However, there were problems related with the workability of the mixes at such high asphalt content, which were solved by determining the compaction and mixing temperature for the mixes. Finally, the fracture testing was done using Superpave IDT testing and the fracture properties were evaluated using the framework of HMA Fracture Mechanics. The issues concerning the IDT testing on the friction course were also identified and dealt with. 1.3.2 Georgia PEM Georgia PEM is a kind of Porous European Mix adopted by Georgia DOT. The primary difference between normal FC-5 and Georgia PEM is the use of polymer modified asphalt and addition of mineral fiber to the mix. Georgia PEM mixes were prepared using the same mix design as used by Georgia DOT. However, Superpave Gyratory Compactor was used instead of Marshall blows for compaction. Both unaged and aged mixes were evaluated for their fracture performance using IDT testing and the results were compared with FC-5 and Novachip.

PAGE 18

CHAPTER 2 LITERATURE REVIEW An extensive literature review was done covering the mix design, construction and performance of various kinds of friction courses. The main focus of literature review was on 1) various mix design approaches for thick open graded and bonded friction courses, 2) materials, equipment and construction guidelines used by the different states in US. The literature review also covered set of guidelines and protocols for producing representative laboratory mixtures for evaluation and testing, as well as guidelines for the construction of bonded friction courses (Novachip). 2.1 Novachip Novachip is an ultra thin friction course whose primary objective is to restore the skid resistance and surface impermeability. The few other advantages are excellent adhesion, reduced rolling noise, reshaping of existing pavements. 2.1.1 Novachip Description Novachip consists of a layer of hot precoated aggregate over a binder spray application. The tack coat is generally a polymer modified, emulsified asphalt (usually a latex or elastomer modified emulsion. Such a coating offers strong bonding between the chippings. Thus due to the immediate application of the binder, chippings are perfectly held in position and whip-off is totally eliminated. The hot mix material is a gap graded mixture that includes large proportion (70-80%) of single sized crushed aggregate, bound with mastic composed of sand, filler and binder. The binder content varies usually from 5-6 percent. The course thickness varies 5

PAGE 19

6 from 10-20 mm depending on maximum size of stone. Layer thickness is usually 1.5 times the diameter of the largest stone. Novachip is placed with a specially designed paving machine that combines the function of asphalt distributor and a lay down machine. The paver applies the tack coat and the hot mixture in a single pass. This heavy application of tack helps to ensure adhesion of the friction course to the underlying pavement and to prevent the possibility surface water from permeating into the pavement. The operation of the paving equipment has been described in detail by Colwill et al. (1) and Serfass et al. (18). 2.1.2 Paving Equipment The operation involved in the Novachip process is as follows Collection of mixture from the transport truck Storage of mixture Storage of sufficient tack emulsion for at least 3 hrs of operation Distribution of tack coat with servo controlled application rate Immediate covering of tack with the mixture Smoothing of applied mixture The Novachip machine includes the following components: A receiving hopper for precoated chippings with self locking hook for the supplying truck A scraper-type conveyer A chipping storage chamber with appropriate thermal insulation and a total capacity of 5 m 3 Several binder tanks, thermally insulated, with a capacity of 12 m 3 (more than half day work) A conveyer transferring chippings to the screed unit A variable-width spray bar A variable width heating screed unit The design of spray bar includes wide-angle nozzles whose delivery depends on the road speed of the machine. The friction coarse is applied at high speed, > 10m/min and reaches 20-25 m/min.

PAGE 20

7 2.1.3 Post Construction Testing Performance evaluation of NOVACHIP has been typically in terms of measurement of skid resistance, surface macrostructure, surface roughness, and ride quality data and noise level. The procedure for above tests are described as follows: 2.1.3.1 Surface roughness International roughness index The international Roughness Index (IRI) measures pavement roughness in terms of the number of inches per mile that a laser, mounted in a specialized van, jumps as it is driven across the interstate and expressway system (10). The lower the IRI number, the smoother the ride. The rating system scores a roadway from the following criteria: Table 2.1 International Roughness Index FHWA Highway Statistics Categories (Inches per mile) Interstate Routes National Highway System (NHS) Non Interstate Routes Non-NHS Traffic Routes & Other Routes with ADT>= 2000 All Other Routes <60 Excellent Excellent Excellent Excellent 60-94 Good Good Good 95-119 Fair Fair Good 120-144 Fair 145-170 Fair 171-194 Poor Poor 195-220 Poor >220 Poor The test is conducted in accordance with ASTM E-950 test method. 2.1.3.2 Skid friction Pavement skid friction is tested in accordance with ASTM E-274 test method (locked wheel skid trailer) (10). The classes of skid numbers are as follows

PAGE 21

8 Table 2.2 Skid Number Skid Number Description 50-100 Very Good 40-49 Good 30-39 Fair 20-29 Poor 1--19 Very Poor 2.1.3.3 Surface macrostructure Pavement friction depends on both microstructure and macrostructure. Microstructure refers to detailed surface characteristics of the material. Good microstructure will provide effective contact area between the tire and the aggregate on the road surface. Macrostructure refers to the general coarseness of the surface, which facilitates the drainage of water from the surface. Pavement macrostructure is defined as the deviation of the pavement from a true planar surface and the average texture depth between the bottom of the pavement surface and the top of surface aggregates. Surface macrostructure depth is measured by Sand Patch Depth (SPD) volumetric technique method in accordance with ASTM E 965-87 (10). 2.1.3.4 Ride quality data Texas DOT uses an instrument called SIometer to measure the ride quality of the pavement surface (3,4). The ride quality essentially indicated the smoothness or irregularities and the ruts in the pavements. A SIometer has an accelerometer processing computer and a data storage computer mounted in one vehicle. The data is converted into a ride score based on a user panel rating than ranges from 0.1 (very rough) to 5 (very smooth). The ride score classifications are shown in the table as follows:

PAGE 22

9 Table 2.3 Ride Quality Ride Score Description 4.0-5.0 Very Smooth 3.0-3.9 Smooth 2.0-2.9 Medium Rough 1.0-1.9 Rough 0.1-0.9 Very rough 2.1.3.5 Rolling noise Rolling noise measurement can be made using a French-German method i.e using three different vehicles with different types of tires at various speeds (18). This method establishes a correlation between acoustical pressure (in dB) and vehicle speed. 2.1.4 NOVACHIP Mix Design The NOVACHIP mix design is based on the design developed by the FHWA office of research (19). The material requirement, aggregate gradations are established and the bulk and apparent specific gravity of coarse and fine aggregate are determined. The specific gravity of asphalt cement is also obtained. After these preliminary steps, surface capacity (K c ) of the coarse aggregate is calculated and it is used to then calculate the design asphalt content. 2.1.4.1 Gradation There are three kinds of gradation band based on the nominal maximum size (9,10). These design gradation bands for the various nominal maximum sizes are as follows 2.1.4.2 Asphalt content determination The surface capacity K c is determined by the Centrifuge Kerosene Equivalent (C.K.E) test based on FHWA design procedure (20). The K c value is a measure of relative roughness and degree of porosity of the aggregate and is used in an experienced based formula to calculate the design asphalt content.

PAGE 23

10 Table 2.4 JMF Gradation Range Sive Size (mm) Sive Size^0.45 Percentage Passing A (6.3 mm) B (9.5 mm) C (12.5 mm) Max Min Max Min Max Min 19 3.76 100 100 12.5 3.12 100 100 100 85 9.5 2.75 100 100 100 75 90 70 6.3 2.29 100 75 45 30 45 30 4.75 2.02 60 40 37 24 40 24 2.36 1.47 24 20 26 21 25 21 1.18 1.08 20 15 23 15 25 18 0.6 0.79 15 10 15 12 20 12 0.3 0.58 12 8 14 8 16 8 0.15 0.43 10 7 10 5 10 5 0.075 0.31 7 5 7 5 7 4 Asphalt Content % 5.3 % Min 6 % Max Equipment. Pans, 115 mm 25 mm deep. Hot plate or oven capable of 110 C 5 C ( 230 F 9 F) Beaker, glass 1500 ml Balance, 500 g capacity, Metal funnels, top dia 89 mm, height 114 mm, Orifice 13 mm with piece of 2mm sieve soldered to the bottom of the opening Oil, S.A.E No. 10 lubricating Surface Capacity (K c ) Test for Coarse Aggregate. 1. Quarter out 105 g of aggregate representative of material passing through 9.5 mm sieve and retained on 4.75 mm (No. 4) sieve. 2. Dry sample on hotplate or in110 C 5 C (230 F 9 F) oven to a constant mass and allow to cool 3. Weigh out approximately 100 g sample to nearest 0.1 g and place it in a funnel 4. Completely immerse sample in S.A.E No. 10 lubricating oil for 5 minutes. 5. Drain sample for 2 minutes 6. Place funnel containing sample in 60 C (140 F) oven for 15 minutes of additional draining 7. Pour sample from funnel into a tarred pan, cool, reweigh sample to nearest 0.1 g. Determine the amount of oil retained as percent of dry aggregate mass

PAGE 24

11 8. Use the chart as shown in Fig. 2.1. If the apparent specific gravity for the fraction is greater than 2.70or less than 2.60 apply correction to percentage oil retained using the formula as shown in Fig. 2.1. Design Asphalt Content. 1. If the apparent specific gravity of the coarse aggregate fraction is 2.6 or more but not greater than 2.7 then 2. AC = 2 Kc + 4.0 3. If the apparent specific gravity of the coarse aggregate fraction is greater than 2.70 or less than 2.60 the 4. AC = (2 Kc + 4.0) 2.65/ Sf 5. AC = Design asphalt content, present by mass of aggregate 6. Sf = Apparent specific gravity of coarse aggregate Thus this the optimum asphalt content used for final design. 2.1.5 Construction Requirements The construction details of NOVACHIP as described in Mississippi DOT specification (19) is as follows: 2.1.5.1 Surface Treatment Paving Machine Screed unit The machine shall be equipped with a heated screed. It shall produce finished surface meeting the requirements of the typical cross section. Extensions added to the screed shall be provided with the same heating capability as the main screed unit, except for use on variable width tapered and/ or as approved by the Engineer. The screed with extensions if necessary shall be of such width as to pave an entire lane in a single pass. Asphalt distribution system. A metered mechanical pressure sprayer shall be provided on the machine to accurately apply and monitor the rate of application of the tack coat. The rate shall be uniform across the entire paving width. It shall be applied at a temperature of 160 F 20 F.

PAGE 25

12 Fig.2.1 Surface Constant Kc vs. Percentage Oil Retained Application can be immediately in front of the screed unit. The application rate shall be 0.22 0.05-gallons/sq. yards, unless otherwise directed by the Engineer.The application rate shall be verified by the carpet tile test before the work commences. At the end of each workday, a check shall be made to determine the quantities of tack coat used. The check shall be made by means of calibrated load cells on the machine Tractor unit. The tractor unit shall be equipped with a hydraulic hitch sufficient in design and capacity to maintain contact between the rear wheels of the hauling equipment and the pusher rollers of the finishing machine while the paving is being unloaded. The machine shall support no portion of the weight of the hauling equipment, other than the connection. No vibrations or other options of the hauling equipment, which

PAGE 26

13 could have a detrimental effect on the riding quality of the completed pavement, shall be transmitted to the machine. The use of any vehicle which requires dumping directly into the finishing machine and which the finishing machine cannot push or propel to obtain the desired lines and grades without resorting to hand finishing will not be allowed. 2.1.5.2 Rollers Steel wheel rollers shall meet be rated at 10 tons and may be three wheel type but the tandem type is preferred. 2.1.5.3 Straightedges and templates When directed by the engineer, the contractor shall provide acceptable 10-foot straightedges for surface testing. 2.1.5.4 Weather limitations The tack coat and the paving mixture shall be placed only when the temperature of the surface to be overlaid is no less than 50 F and rising, but shall not be placed when the air temperature is below 60 F and falling. It is further understood that the tack coat or paving mixture shall be placed only when the humidity, general weather conditions and moisture conditions of the pavement surface are suitable in the opinion of the engineer. 2.1.5.5 Tack coat Before the tack coat and the paving mixture are applied, the surface upon which the tack coat is to be placed shall be cleaned thoroughly to the satisfaction of the engineer. The surface shall be given a uniform application of tack coat in using asphaltic materials as specified approximately two seconds prior to the placement of the paving mixture.

PAGE 27

14 2.1.5.6 Hauling equipment It shall be required that the truck beds be covered while transporting the paving mixture unless otherwise directed by the engineer. At the discretion of the engineer, the truck beds may be insulated. 2.1.5.7 Spreading and finishing The paving mixture shall be delivered at a temperature between 290 F and 330 F unless otherwise directed by the engineer. The paving mixture shall be dumped directly into the surface treatment paving machine and spread on the tack coated surface within two seconds of tack coat having been applied The paving mixture shall be spread to the depth and width that will provide the specified compacted thickness, grade and cross section. Placing of the paving mixture shall be as continuous as possible. The finished surface shall be smooth and of uniform texture and density. 2.1.5.7 Compaction Immediately following placement of the paving mixture, the surface shall be rolled to accomplish a good seating without excessive breakage of the aggregate. A minimum of three passes with steel wheel rollers is required. The compaction shall be accomplished prior to the paving mixture cooling below 180 F. The operation of the rollers shall cause no displacement or showing of the paving mixture. If the displacement occurs, it shall be corrected to the satisfaction of the engineer. Sprinkling of the fresh mat shall be required, when directed by the engineer, to expedite opening the roadway to the traffic. Sprinkling can be with water or limewater solution.

PAGE 28

15 2.1.5.8 Method of measurement Paver laid surface treatment, complete in place and accepted will be measured by the ton. Tack coat will be measured by gallons. 2.2 Porous European Mix (PEM) PEM also known as porous asphalt is a coarse graded mix with around 4-5 % binder and 3% of filler. The binder is typically polymer modified. PEM is designed to contain about 20 % air voids to make it very porous (5,6,7). The differences between PEM and conventional OGFC are as follows: European mixtures generally allow more gap-graded mixtures as compared to the North American mixtures. However, the GDOT specification for PEM is similar to the European mixtures. All European agencies specify minimum air voids. Air voids in OGFC tend to be significantly lower than PEM (around 14 %). European agencies use modified asphalt binders since modified binders are less susceptible to draindown, during both construction and service. European agencies demand higher standards for aggregate than US agencies. LA abrasion values are specified from 12-21 %. For OGFC it between 35-40 % European agencies specify minimum asphalt content based on durability test (Cantabro test), which is performed on compacted specimen. A maximum limit is based on air voids. In contrast to this, for OGFC asphalt content is selected based on FHWA method. The asphalt content in US mixtures typically varies from upper five to mid six percent ranges It has been observed that use of PEM has led to hydroplaning and glare reduction, noise reduction and increase in skid friction. However, the major disadvantages include lack of strength and clogging of the pores. The latter problem can be solved using multiple layers with air voids increasing as we go down. 2.2.1 Advantages of Porous Asphalt 2.2.1.1 Hydroplaning and Glare Reduction On of the major hazards while driving in rain is aquaplaning. A layer of water builds between the surface and tire thus causing the vehicle to literally float. As a result,

PAGE 29

16 it becomes difficult to steer and apply the brakes. With porous asphalt it is designed, to have interconnected air voids for high permeability and higher macrostructure, which solves the problem of aquaplaning to a great extent. With porous asphalt, wet weather visibility is significantly improved since there is little or no free surface water. Usually on unlit roads, the nighttime visibility decreases due to the water layer on the road surface, which hampers with the retro reflecting devices. In addition, due to the water logging, mirror reflections are caused by the headlights of the oncoming vehicles. However, porous asphalt is effective in these conditions in reducing the glare. 2.2.1.2 Noise Reduction One of the benefits of porous asphalt mix is the significant noise reduction compared to the denser mixes due to good absorption potential. Measurements in most European countries have shown that by using porous asphalt noise levels can be reduced by approximately 3 dB(A) as compared to the conventional dense asphalt concrete surfacing (11). These figures apply to passenger car vehicle traveling at 80 Km/hr in dry conditions. The noise level is influenced by aggregate size, size distribution, permeability and the condition of the layer. It should be noted that porous asphalt not be used in high traffic low speed roads since it has been observed that after few years (typically 3 years) of service all noise reduction benefits are lost because the surface voids get clogged and the mixture in turn becomes dense graded. However, at high speed, the hydraulic action flushes the dirt from the pavement voids and reduces clogging. In addition, the cleaning action would involve spraying or vacuuming. The vehicle noise is generated by different mechanisms. At high speed, the vibrations in the tire structure and air pumping in the cavities underneath the tire cause

PAGE 30

17 tire noise. The tread blocks slip on the road surface mostly at the edges of the contact area where the vertical forces are lower and the tire deforming as the weight is applied or removed. When the tire rolls forward and the tread blocks snap back to their unloaded shape it generates vibrations on the tire surface thus generating large noise. Thus, movement of air in the cavities of tire treads cause air pumping. As the tire rolls forward the compressed air tries to escape to the sides. As the tread block is lifted vacuum is created in the tread cavities and the air rushes to fill the voids. If the pavement texture is smooth, there is less opportunity for the air to leak from underneath the tread block and the tire noise is accentuated. 2.2.1.3 Skid Friction Rain may reduce skin friction of road considerably even when no aquaplaning takes place. Porous asphalt counteracts this effect and even at high speeds the skin friction with porous asphalt is high on wet roads (1,2). Finally, on dry roads due porous asphalt gives exceptionally good skidding properties at higher speeds where macrostructure is very important. 2.2.2 Disadvantages of Porous Asphalt 2.2.2.1 Strength Porous asphalt does not contribute to the overall structural integrity of the pavement due to high air voids. In Belgium, it was observed that the moduli of porous asphalt manufactured with asphalt 80/100, was 73-79 percent of that of wearing course in conventional asphalt concrete (5). In Netherlands, the structural behavior of porous asphalt is assessed using the Department of Public Works standard multi layer elastic layer design The model assumes that the fatigue resistance of the road pavements is determined by the lower part

PAGE 31

18 of the structure, which ignores the possibility of fatigue cracks being developed in the upper part of the structure. Studies have indicated that three aspects need to be specifically addressed to evaluate the bearing capacity of the porous asphalt, which are as follows: 2.2.2.2 Initial Stiffness Modulus Using fatigue test initial E-modulus is determined which is used in the elastic design model to estimate the contribution of porous asphalt to bearing capacity of the pavement structure. From the study conducted, it was observed that initial effective contribution was around 80 to 90 percent of that attainable with gravel asphalt concrete, depending on thickness of structure. Thus, these mixes tend to have lower stiffness due to the nature of their structure. 2.2.2.3 Aging and Stripping As a result of open structure of porous asphalt, the binder is likely to undergo oxidation to undergo accelerated aging which in turn increases the stiffness of binder considerably. Also water ingress will lead ton stripping of the lower part of the surface layer, which affects the cohesive properties as well as the adhesion to the underlying base course thus impairing the load transfer characteristic of the structure. Based on the studies conducted it was observed that the weighted effective contribution was 35-40 percent of that achieved by dense asphalt concrete. 2.2.2.4 Temperature The suction and pumping action of tires passing over porous asphalt layers, coupled with wind motion, will promote a continuous circulation of air within pores. Consequently, the temperature in porous asphalt wearing course is likely to remain closer to the prevailing air temperature than with the closed surfacing materials. Analysis of

PAGE 32

19 study on temperature gradient on porous asphalt and dense asphalt concrete showed that the weighted average temperature over a year was found to be about 1C lower in pavements surfaced with porous asphalt than inn comparable structures with dense asphalt concrete wearing course. Thus, the stiffness of porous asphalt is less affected by warm weather. Thus the combined of the above three factors implies that porous asphalt can be expected to contribute to about 50 percent of the equivalent bearing capacity achievable with dense asphalt concrete. 2.2.2.5 Clogging The service life of porous asphalt is generally less due to premature clogging of the voids, which leads to ineffective drainage of the surface water. Also when the surface pores become plugged, a pavement might fail by asphalt being stripped from the aggregate surface. However, in Netherlands, they use a new technique of using two layers of porous asphalt to counter this problem (6). The surface layer uses aggregate that is 4 to 8 mm in size. Directly underneath is another porous asphalt layer containing 11 to 16 mm sized aggregate. The surface layer allows water and sound to penetrate to larger chamber of voids in the lower layer. The surface layer has smaller voids to prevent larger materials from clogging the surface voids. The smaller debris, which enters the surface voids, can be suctioned out by the hydraulic action of the tires on the pavements. 2.2.3 Performance Related Laboratory Testing 2.2.3.1 Resistance to Plastic Deformation As a part of laboratory study carried out by University of Cantabria and ESM Research Centre (8), Spain, the resistance to deformation was found out using wheel-tracking test at a temperature of 60C. The percentage of binder used ranged from 4 to 4.5

PAGE 33

20 for the porous asphalt mixes. It was observed that modified asphalt offered greater resistance to plastic deformation than that of ordinary asphalt. Thus, use of polymeric asphalt diminishes the effect of post compaction by the traffic, which is sometimes observed in the porous mixes 2.2.3.2 Resistance to Indirect Tension The effect of binder on improving resistance to traction was studied in Spain through IDT testing on Marshall samples as a part of research to study the effect of special binders on porous mixes (16). The test temperatures chosen were 5 and 45C and the application rate of 50.8 mm/min. The samples are compacted with energy of 50 Marshall blows per face. The test results showed that the performance of mix made with polymeric asphalt was better than with ordinary asphalt. The difference was more prominent at 45C than at 5C. 2.2.3.3 Resistance to Disintegration Disintegration resistance is tested in the laboratory through the Cantabro test of wear loss, consisting of testing Marshall samples in the LA abrasion machine and obtaining the weight loss after 300 drum revolution (9). This test is a standard test for porous asphalt prevalent throughout Europe. The test is used to predict the maximum and minimum asphalt content, however the specifications and the test temperature vary with countries. 2.2.3.4 Adhesiveness Resistance to stripping is influenced by the adhesiveness between the aggregate and the binder. For studying resistance to stripping, in Spain and UK, Cantabro test is used to determine the loss in the test sample that has been submerged in water at a controlled

PAGE 34

21 temperature (around 49C) for 4 days (9). It is observed that modified asphalt has grater adhesiveness than ordinary asphalt though the losses are higher when immersed in water. 2.2.3.5 Drainage Test Basket Drainage test (2) has been used in Belgium and Spain for evaluating the draindown of the porous asphalt mixes. The procedure is as follows: The mix is manufactured and compacted in Duriez mold under a pressure of 30 bars The molds are then laid on a grid and the set is placed on an oven at 180C for 7 hours, these severe conditions are chosen to simulate occasional cases when the asphalt is draining through the aggregates The asphalt drained through the mix to the grid is recovered and the loss of asphalt is calculated with respect to the initial binder content. 2.2.4 Mix Design Approach The design of porous asphalt (6,9,16) is based on A minimum binder content to ensure resistance against particle losses and thick film on the aggregates A maximum binder content to avoid binder runoff and have a good drainability in the mix Using the maximum abrasion value, a minimum amount of binder is fixed. The initial selection of the binder type is influenced by the aggregate source and the amount of binder that needs to be carried. The purpose of using modified binders is to improve the resistance against particle losses with very open mixtures through higher cohesion and get a longer durability through thicker films of binder because of higher viscosity. A reduction in thermal susceptibility is also sought to get higher consistencies with high temperature and more flexibility wit low temperatures. The layer thickness is typically around 4 cm (11). The mix design approach followed by various European countries is as follows:

PAGE 35

22 2.2.4.1 British Design The mix design approach followed in Britain (6) is as follows: Two types of porous asphalt mixes are used; one is 10 mm nominal maximum size and the other is 20 mm nominal maximum size. The 10 mm is more like that of other European countries. 20 mm is considerably larger. The gradation is as shown in Fig.2.2 The aggregates are specified to provide a hard and durable rock. The LA value for aggregates must be low, less than 12% or 18%. The aggregates must have high polishing resistance and all the aggregates should have at least two fractured faces. Maximum flakiness index of 25 to control aggregate shape Both modified and unmodified asphalt binder used. The binder is modified with Styrene-butadiene-styrene (SBS) or ethylene-vinyl acetate (EVA) modifier Specimen compacted using 50 Marshall Blows. Durability is not tested directly but is controlled through selection of binder grade and minimum asphalt content Minimum asphalt content is 4.5 % and minimum air voids is 20% 2.2.4.2 Spanish Design The mix design approach followed in Spain (6) is as follows: There are two gradations of porous asphalt used in Spain. Both are 12.5 mm nominal maximum size. The P12 has larger gap gradation between 12.5 and 10 mm sieve. The PA12 has larger gap gradation between 10 and 5 mm sieve. The gradation is as shown in Fig.2.3 A maximum LA abrasion value of 20 % for the aggregates A maximum flakiness index of 25 % Modified asphalt binder used (for high traffic and hot climate 60/70 penetration asphalt is used. For low traffic and cool temperature, 80/100-penetration asphalt is used). The modifier used is SBS or EVA Specimen compacted using 50 Marshall Blows Minimum air voids of 20% selected using Cantabro test with a maximum weight loss of 25 % at 25C Maximum asphalt content is based on the air voids of the compacted specimen. Typical asphalt content is around 4.5%

PAGE 36

23 2.2.4.3Italian Design The mix design approach followed in Italy (6) is as follows They use just one gradation, a 16 mm nominal maximum size aggregate. All aggregate are crushed with no natural sand allowed. The gradation is as shown in Fig.2.4. A maximum LA abrasion of 16% for the aggregates A modified binder is used. 6-8 percent is added to 80/100-penetration asphalt. Mix temperature is 190-200C. They use SBS as modifier. Minimum asphalt content is 4.5 percent based on Cantabro test. The allowable asphalt content ranges from 4-6 percent. A maximum weight loss of 25 percent is allowed at 20C. Air voids vary from 18-23 percent Moisture damage is evaluated during drainage using Cantabro test. Maximum loss allowed is 30% 2.2.4.4 Belgium Design The mix design consist of first determining the voids in the coarse aggregate and then measuring with various binder content the voids and the percentage of wear estimated using Cantabro test Aggregates should contain more than 80 percent stones greater than 2mm in diameter. The proportion of sand (.08 2 mm) should be around 12 percent and rest should be filler A gap grading is to be obtained by omitting 2/7 or 2/10 mm fraction from 0/14 mm mixture Air voids is around 16-18 percent Asphalt content is around 4-5 percent (unmodified). For modified binders, it should be around 5.5 to 6.5 percent. 2.3 Georgia PEM As already mentioned, OGFC mixes are typically gap-graded and contain high percentages of single sized coarse aggregate. They typically have high AC content, a thick AC film, low percentage of material passing the 0.0075 mm sieve and high volume of air voids (18-22%). The material composition of Georgia PEM is as described below. 2.3.1 Material Selection 2.3.1.1 Composition The JMF consist of mainly coarse aggregate, typically granite, with small amount of fines. The binder used should be very stiff such as PG 76-22 made with polymers (typically SB or SBS). The addition of fibers is desirable since it reduces drain down.

PAGE 37

24 0204060801001200.1110100Size (mm)% Passing (A) 0204060801001200.1110100Size (mm)% Passing (B) Fig. 2.2 British Gradations A) 20 mm B) 10 mm

PAGE 38

25 0204060801001200.1110100Size (mm)% Passing (A) 0204060801001200.1110100Size (mm)% Passing (B) Fig. 2.3 Spanish Gradation A) P12 B) PA12

PAGE 39

26 0204060801001200.1110100Size (mm)% Passing Fig 2.4 Italian Gradation Band 2.3.1.2 Gradation The job mix formula is used for the Georgia PEM (21) is as shown in Table 2.5 Table 2.5 Georgia PEM Gradation Sieve Size (mm) Size ^ 0.45 12.5 mm PEM (% Passing) Max Min 19 3.76 100 100 12.5 3.12 100 80 9.5 2.75 60 35 4.75 2.02 25 10 2.36 1.47 10 5 0.075 0.31 4 1

PAGE 40

27 The 0.45 power chart is as follows 0204060801001200.000.501.001.502.002.503.003.504.00Size ^0.45% Passing Fig. 2.5 Georgia PEM Gradation Band 2.3.1.3 Aggregate Specification The specifications for basic aggregate properties are as follows (21) Table 2.6 Aggregate Specifications for Georgia PEM Parameter Requirement LA abrasion Loss (%) <50 Soundness Loss (%) <15 Flat ans Elongated Particles (5:1 ratio) <10 Mica Schist Ratio <10 Note that only silica rich aggregates can be use (e.g. granite). Carbonaterich aggregates are excluded. Soundness loss is measured using magnesium sulfate. Typical loss is less than 2%. 2.3.1.4 Polymer Modified Asphalt cement GDOT primarily uses two polymers, styrene butadiene (SB) and styrene butadiene styrene (SBS), to modify cements used in OGFC mixes. The main improvements after

PAGE 41

28 the inclusion of polymers have been 1) increase in binder stiffness by 8-10 times 2) the softening point of AC has increased by approximately 44F and 3) the AC is more ductile and flexible than unmodified AC. Due to greater viscosity of the polymer blend, temperature requirements in the design procedure have been increased to 325F. In addition, a phase angle requirement of less than 75 has been added to help ensure that polymer modification is used to meet binder grade requirement. The base asphalt cement is typically modified with 4.0-4.5 % polymer by weight of AC. 2.3.1.5 Mineral Fibers Fibers are used in OGFC to stabilize the AC film surrounding aggregate particles in order to reduce AC draindown during production and placement. GDOT uses mineral fibers in OGFC at a dosage rate of 0.4 % by weight of total mix. Thus, the final design requirements have been tabulated in the following table: Table 2.7 Design requirements for Georgia PEM Parameter Requirement Binder Content 5.5 7.0 Polymer SB or SBS (%) 4-4.5 Air Voids (%) 20 24 Drain Down < 0.3 PG 76-22 Fibers (%) 0.2 0.4 Anti Stripping (lime) (%) 1.00 2.3.2 Georgia PEM Mix Design Procedure (GDT 114) Scope The design for the Georgia PEM consists of four steps. The first step is to conduct AASHTO T-245 to determine asphalt cement content then, secondly to determine

PAGE 42

29 optimum asphalt content. The third step to perform GDT-127 and final step is performing GDT56. Apparatus 1. 13 metal pie pans 2. Oven capable of maintaining 250 F 3.5 F 3. Oven capable of maintaining 140 F 3.5 F 4. Beaker glass, 500 ml 5. Glass funnels, top dia = 3.5 in; height = 4.5 in; orifice = 0.5 in with piece of No. 10 sieve positioned at top of funnel neck. Cork stopper to fit the outlet of funnel neck 6. Oil S.A.E No. 10 lubricating 7. Drain Down Equipment as specified in GDT-127 8. Marshall Design equipment as specified in AASHTO T245 9. Equipment as specified in GDT 56 10. Balance 5000 gms, 0.1 gm accuracy Step 1 Surface Capacity (Kc) Quarter out 105 g of aggregate representative of material passing through 9.5 mm sieve and retained on 4.75 mm (No. 4) sieve. Dry sample 250F 3.5F oven to a constant mass and allow to cool Weigh out approximately 100 g sample to nearest 0.1 g and place it in a funnel Completely immerse sample in S.A.E No. 10 lubricating oil for 5 minutes by plugging funnel outlet with cork stopper Drain sample for 2 minutes Place funnel containing sample in 60C (140F) oven for 15 minutes of additional draining Pour sample from funnel into a tarred pan, cool, reweigh sample to nearest 0.1 g. Determine the amount of oil retained as percent of dry aggregate mass Use the chart as shown in Fig.6.2. If the apparent specific gravity for the fraction is greater than 2.70or less than 2.60 apply correction to percentage oil retained using the formula shown in Fig.2.1. Determine the required asphalt using the following formula % AC = 2.0 (Kc) + 3.5 (No correction applied for viscosity) Step 2 Modified Marshall Design and Optimum AC 1. Heat the coarse aggregate to 350F 3.5F, heat the mould to 300F 3.5 F and heat the AC to 330 F 3.5 F 2. Mix aggregate with asphalt at three asphalt contents in 0.5 % interval nearest to the optimum asphalt content establishes in step 1. The three specimens should be compacted at the nearest 0.5% interval to the optimum and three specimens each at 0.5% above and below the mid interval.

PAGE 43

30 3. After mixing, return to oven if necessary and when 320F 3.5F compact using 25 blows on each side 4. When compacted, cool to the room temperature before removing from the mold 5. Bulk Specific Gravity: 6. Determine the density of a regular shaped specimen of compacted mix from its dry mass (in grams) and its volume in cubic centimeters obtained from its dimensions for height and radius. Convert the density to the bulk specific gravity by dividing by 0.99707 g/cc, the density of water at 25C 7. Bulk Sp.Gr = W / ( r2h/ 0.99707) = Weight (gms) 0.0048417/Height (in) W = Weight of specimen in grams R = radius in cm H = height in cm 8. Calculate percent air voids, VMA and voids filled with asphalt based on aggregate specific gravity 9. Plot VMA curve versus AC content 10. Select the optimum asphalt content at the lowest point on VMA curve Step 3 Drain-Down Test Perform the drain test in accordance with the GDT 127 (Method for determining Drain Down characteristics in Uncompacted Bituminous Mixtures). A mix with an optimum AC content as calculated above is placed in a wired basket having 6.4 mm (1/4 inch) mesh openings and heated 14C (25F) above the normal production temperature (typically around 350F) for one hour. The amount of cement, which drains from the basket, is measured. If the sample fails to meet the requirements of maximum draindown of 0.3%, increase the fiber content by 0.1% and repeat the test. Step 4 Boil Test Perform the boil test according to GDT 56 with complete batch of mix at optimum asphalt content as determined in step 2 above. If the sample treated with hydrated lime fails to maintain 95% coating, a sample shall be tested in which 0.5% liquid anti stripping additive has been used to treat the asphalt cement in addition to the treatment of aggregate with hydrated lime.

PAGE 44

CHAPTER 3 MATERIALS Five test sections were laid on SR-27 in Highlands County, in July 2003, as part of friction course study. The section consisted essentially of three kinds of mixes, viz., FC-5 with Limestone, FC-5 with Granite and Novachip. For Georgia PEM, Nova Scotia-granite was used and the mixes were prepared in the lab. The gradation was exactly same as specified by the Georgia DOT. The details of the material gradation and binder type are described in the following sections. 3.1 FC-5 and Novachip 3.1.1 Aggregates For FC-5, two kinds of aggregates were used: Limestone and Granite. In case of Novachip, Nova Scotia-granite was used. The gradations for these mixtures are as shown in Fig 3.1. Clearly the FC-5 Granite and FC-5 limestone seems to have much coarser gradation that Novachip. The granite was treated with hydrated lime for prevention against stripping 3.1.2 Binder For FC-5, the binder used was AC-30 with 12% ground tire rubber. For Novachip PG76-22, a polymer modified binder, was used. The binder type and binder content used for the field section is as shown in Table 3.2.The novabond is a polymer modified tack coat and is same as SBS modified PG 76-22. 31

PAGE 45

32 Table 3.2 Mixes from the field Type Binder Type Binder Content FC-5 Limestone ARB-12 6.40% FC-5 Limestone With NOVABOND ARB-12 6.40% FC-5 Granite ARB-12 6% FC-5 Granite With NOVABOND ARB-12 6% NOVACHIP PG 76-22 (SBS Modified) 5% 3.2 Georgia PEM 3.2.1 Aggregate and Binder Nova Scotia-granite was used for the mix. Hydrated lime was used as anti stripping agent. The aggregate gradation used is as shown in Fig 3.1 SBS modified PG 76-22 was used and mineral fiber was added at 0.4% of the mix as per the requirements. Typically, if the gradation is within the prescribed limit, the asphalt content is usually around 6% and the air void level is around 20%, which is what is targeted in these mixes.

PAGE 46

33 01020304050607080901000123Sieve Size (raised to 0.45 power) mmPercentage passing FC-5Granite FC-5Limestone Novachip GPEM 0.0750.3000.6001.182.364.759.519.00.15012.5 Fig. 3.1 Gradation of FC-5 Granite, FC-5 Limestone, Novachip and Georgia PEM

PAGE 47

CHAPTER 4 DETERMINATION OF COMPACTION LEVEL AND AGING PROCEDURE FOR FRICTION COURSES It is a well-known fact that during compaction in the field, a stage is reached where the aggregate resistance to compaction increases considerably. In other words, there is a great degree of interlocking between the aggregates. Hence, during compaction in lab it becomes important to identify the stage at which the mix exhibits this interlocking. This point of interlocking is called the Locking Point. This concept was first defined by Vavrik in 1998 for dense graded HMA. The following section focuses on identification of the locking point for the friction courses. 4.1 Compaction of the Friction Course Mix Initially each of the FC-5 limestone, FC-5 granite and NOVACHIP mixes were compacted to 125 gyrations (N design for traffic level 5 as per Superpave criteria) in the Superpave Gyratory Compactor. This was used to study further the compaction level for these mixtures. 4.1.1 Compaction Data The bulk specific gravity of the mixes was obtained from the compaction data. While calculating Gmb dimensional analysis was done, i.e. Gmb est was used for calculation of air voids. This was because the mixtures were so coarse that the bulk specific gravity experiment would not have yielded appropriate value of Gmb. The volumetric properties of the mixes are as follows: 34

PAGE 48

35 Table 4.1 Air Voids Type Gmb Gmm Air Voids FC-5 Granite 2.012 2.336 13.87 FC-5 Limestone 1.958 2.441 19.78 Novachip 2.135 2.474 13.7 The graphs of percentage Gmm Vs N for all the pills are as shown in Fig. 4.1 Fig.4.3. The compaction curve (% Gmm Vs N) follows logarithmic relationship strongly. Hence, statistical regression was done using a logarithmic relation. Thus, the equation of the regression curve was as follows: %Gmm = m ln(Ngyr) + c i.e Ngyr = exp((%Gmm-c)/m) Where m is the slope of the curve at a given gyration and c is a constant. Now, the locking point (Vavrik, 1998) is defined as the first three gyrations that are at the same height preceded by two gyrations at same height (the height is in taken mm up to single decimal place). However based on the above definition, it was found that compaction of FC-5 with limestone did not yield any locking point which meant that the limestone had probably been crushed during compaction. Further, even though we could identify the locking point in case of NOVACHIP and FC-5 with granite, the gyrations seemed to be on the higher side since the air voids had more or less reached a constant value by then. (Refer to Table 4.3 for locking points for the mixtures as per Vavricks definition). Hence, in order to ascertain whether crushing of the aggregates had taken place, the gradations of these compacted mixes were determined.

PAGE 49

36 y = 4.7519Ln(x) + 63.379R2 = 0.996864.0066.0068.0070.0072.0074.0076.0078.0080.0082.0084.0086.0088.00020406080100120140Gyrations% Gmm Fig. 4.1 Compaction Curve for FC-5 Limestone y = 3.1847Ln(x) + 70.997R2 = 0.994464.0066.0068.0070.0072.0074.0076.0078.0080.0082.0084.0086.0088.00020406080100120140Gyrations% Gmm Fig. 4.2 Compaction Curve for NOVACHIP

PAGE 50

37 y = 2.6809Ln(x) + 67.651R2 = 0.993164.0066.0068.0070.0072.0074.0076.0078.0080.0082.0084.0086.0088.00020406080100120140Gyrations% of Gmm Fig. 4.3 Compaction for FC-5 Granite To check for the breakdown of the aggregates, the gradations of the mixes were obtained by reflex extraction and compared with the original (ref Fig. 4.7). Looking at the gradation, it becomes clear that crushing had taken place in case of limestone, though, for FC-5 with Granite and NOVACHIP there seemed to be no significant difference in the gradations. However, the compaction curve indicates that for these mixes there is no significant change in the air voids at higher gyrations. Thus, the locking point should be way below the existing level of gyrations. Hence, instead of looking at the height of compaction, we looked at the rate of change of compaction, which was a better indication of resistance to compaction. Thus, studying the rate of change of slope of the compaction curve, new locking point was identified for these friction course mixes. Again, locking point being the point beyond which the rate of resistance to compaction increases significantly, it implies that at this stage, the rate of change of compaction decreases significantly and this is what has been identified in the following study

PAGE 51

38 Table 4.2 Locking point For the Mixtures Type Locking Point as per Varick( # of Gyrations) FC-5 With Lime No Locking Point FC-5 With granite 76, 83 NOVACHIP 86, 91, 96 (sample 1) 84, 89, 94 (sample 2) 4.1.2 Initial Study Since the focus was on resistance to compaction, the exact nature of the compaction was studied by looking at the rate of change of compaction. Now decrease in the rate of compaction is directly proportional to the increase in resistance to compaction. This was essentially used to identify the point of maximum resistance to compaction. It was observed that the compaction curve became linear beyond a certain gyration. This meant that compaction had reached a stage where no further decrease in rate of compaction was possible and this stage was the stage of maximum resistance to compaction. Hence, the point beyond which the compaction curve became linear was identified and it was observed that beyond 50-60 gyrations the curve more or less became linear in nature. This has been presented in Figures 4.4 through 4.6. Thus, the points from visual identification served as reference values for identifying the locking points for these mixtures. It was observed that the compaction curve followed logarithmic trend. To identify the locking point, the rate of change of slope of compaction curve was used. The stage, at which the rate of change of compaction was insignificant, was essentially the point of maximum resistance to compaction.

PAGE 52

39 FCLime-164.0066.0068.0070.0072.0074.0076.0078.0080.0082.0084.0086.0088.00020406080100120140# of Gyrations% Gmm Point beyond which curve becomes linear Fig. 4.4 Locking Point for FC-5 with limestone by Visual Observation Novachip-264.0066.0068.0070.0072.0074.0076.0078.0080.0082.0084.0086.0088.00020406080100120140# of Gyration% Gmm Point beyond which the curve becomes linear Fig. 4.5 Locking Point for Novachip by Visual Observation

PAGE 53

40 FC-5 G64.0066.0068.0070.0072.0074.0076.0078.0080.0082.0084.0086.0088.00020406080100120140# of Gyrations% of Gmm Point Beyond which curvebecomes linear Fig. 4.6 Locking Point for FC-5 with Granite by Visual Observation Thus, using the logarithmic regression of the compaction data, the rate of change of slope can be obtained as, y = m ln(x) + c Rate of compaction = dy/dx = m/x (at any x=N) Rate of change of slope of compaction curve = d 2 y/dx 2 = -m/ x 2 (at any x =N) Based on the above idea the locking point was identified as the point at which two gyrations at same gradient of slope were preceded by two gyrations at same gradient of slope. The gradient was taken up to four decimal places (as shown in Table 4.3 for FC-5 Granite). The reason this was chosen as locking point was based on the fact the change in air voids was insignificant at this stage and that this trend was consistently observed in all the mixtures. In addition, the compaction level as identified from visual observation was around 50-60. Thus, based on the above study, the locking points for theses mixtures were identified as shown in Table 4.4

PAGE 54

41 Table 4.3 Locking Point Based on Gradient of Slope FC-5 Granite # of Gyrations Gradient of slope 39 0.0018 40 0.0017 41 0.0016 42 0.0015 43 0.0014 44 0.0014 45 0.0013 46 (LP) 0.0013 47 0.0012 48 0.0012 49 0.0011 50 0.0011 Table 4.4 Locking Points of all Mixtures based on Gradient of Slope Mixtures Locking Point FC-5 Limestone 56 FC-5 Granite 46 NOVACHIP 50 Thus based on above concept the locking points for FC-5 with Limestone, FC-5 with Granite and NOVACHIP were 56, 46 and 50 respectively. The specimens were compacted again to these gyrations and extraction of asphalt was done to observe the gradations after compaction. The results of the gradations after extraction are as plotted in Fig. 3.7 (it has been compared to the original gradations and gradations at 125 gyrations). For FC-5 Lime even when the gyrations were reduced to 56 from 125, the same amount of breakdown was observed. This clearly indicated that in case of limestone, the breakdown occurred in the initial stages itself i.e. at very low gyrations. Hence, even if the gyrations were to be further reduced, the breakdown was still going to persist. For

PAGE 55

42 FC-5 with granite and NOVACHIP, the gradation looks nearly the same as that of the original gradation. In addition, the air voids for FC-5 Granite and NOVACHP were around 22 % and 15 % respectively, which is usually the case for these open graded mixtures. Thus, from the above the study it is clear that, though the locking point of each of these mixtures differed slightly from each other, it was around 50 gyrations. This was further corroborated by the study done by NCAT on the compaction levels of friction courses. NCAT suggests 50 gyrations as compaction level for all friction courses Thus based on our study from visual observation and rate of change of compaction, NCAT study for friction course, we believe that 50 gyrations should be the compaction level for friction course mixes. The steps involved in identifying locking point based on the above-discussed concept are as follow: 1. Fitting a Logarithmic curve on the compaction curve we get from the Superpave Gyratory compactor 2. Obtaining the gradient of the slope of the compaction curve by taking the double derivative of the equation of the regressed curve 3. The locking point is identified as the point at which two gyrations at same gradient of slope were preceded by two gyrations at same gradient of slope. This is close to 50 gyrations for friction courses Finally, based on compaction up to 50 gyrations, the air voids obtained for these mixtures were as follows: Table 4.5 Air Voids for 50 Gyrations Type Gmb Gmm Air Voids FC-5 Limestone 2.012 2.336 17.60 FC-5 Granite 1.958 2.441 21.52 Novachip 2.135 2.474 15.57

PAGE 56

43 0204060801000123Sieve Size (raised to 0.45 power) mmPercentage passing Original 125Gyrations 46Gyrations 0.07 5 0.30 0 0.60 0 1.1 8 2.3 6 4.7 5 9.519. 0 0.15 0 12. 5 (A) 0204060801000123Sieve Size (raised to 0.45 power) mmPercentage passing Original 125Gyrations 56Gyrations 0.0750.3000.6001.182.364.759.519.00.15012.5 (B) 0204060801000123Sieve Size (raised to 0.45 power) mmPercentage passing Original 125Gyrations 50 Gyrations 0.0750.3000.6001.182.364.759.519.00.15012.5 (C) Fig. 4.7 Gradations after Extraction for A) FC-5 Granite B) FC-5 Limestone C) Novachip

PAGE 57

44 4.2 Asphalt Content Determination The asphalt contents for FC-5 with limestone, Granite and NOVACHIP were 6%, 6.4 % and 5% respectively. For the friction courses, when laid with tack coat, the interface between the fiction course and the underlying layer becomes nearly saturated. The tack coat is heavily polymer modified and essentially contains PG 76-22 binder modified with SBS. Hence, to replicate this field condition in the lab it was essential to nearly saturate the existing mixes with PG 76-22. This tack coat is primarily used in NOVACHIP, but for the field project, the very same tack coat was used even for FC-5 with limestone and Granite. The volumetric calculation for determining weight of asphalt to be added was based on maintaining same VMA and reaching the target air void level. The derivation as shown below: Gmm1 = Initial theoretical maximum specific gravity Gmb1= initial bulk specific gravity of the mixes Gmb2= bulk specific gravity after addition of Pg76-22 W1= Total weight of mix before addition of PG76-22 Vt = Total volume of mix V1= volume of aggregate and asphalt before addition of PG76-22 Vair1 = volume of air voids before asphalt addition Vair2 = volume of air voids finally after adding PG76-22 a2 = Final Percentage of Air Voids Desired Was = weight of PG76-22 added Thus Gmb1= W1/(V1+ Vair1)

PAGE 58

45 V1+ Vair1 = volume of compacted specimen =Vt Vt = W1/Gmb1......................................................................................(1) V1= W1/Gmm1 Now again in this mix we add additional PG 76-22. The total volume Vt remains the same but the percentage distribution of air voids and asphalt content changes. Thus, the Gmm of the mix changes Thus Gmm2 = final maximum theoretical specific gravity = (W1+Was)/(V1+ Vas) Now Gmb2 = (1-a2/100) *Gmm2 = (1-a2/100) (W1+Was)/(V1+ Vas).....................................(2) Where Vas = volume of asphalt added = Was/Gb And Gb= specific gravity of PG76-22 = 1.028 Also Gmb2 = (W1+Was)/Vt (the final volume is same).............................(3) Thus equating (2) and (3) and substituting (1) we can get the amount of asphalt to be added Was = ((1-a2/100)*Vt-V1)*Gb 4.3 Mixing and Compaction As already mentioned, to replicate the interface consisting of tack coat, it was decided that we add PG 76-22 since the tack coat contained the same binder. Hence was decided that the air voids in the saturated samples be reduced to 2 %. The mixing and compaction temperature for all the three mixes was 320F.

PAGE 59

46 During mixing, it was observed that amount of drain down was excessively high and it was difficult to work with these mixes at such high asphalt content. Further, compaction was virtually impossible because of such high drain down. Hence, it to was decided that the asphalt added should be such that air voids were reduced by 50% in case of Novachip and 25 % in case of FC-5 with granite and limestone. Again, initially the mixing and compaction temperature was kept as 320F. However, the amount of drain was still significant. Then it was decided that both mixing and compaction temperature be brought down to 255F. However, because of polymer-modified asphalt, mixing became very difficult since the asphalt was very viscous at such a low mixing temperature. Finally, a mixing temperature of 320F and compaction temperature of 255F was adopted and workability of these mixtures improved significantly. However, since the asphalt content was still very high, the loss in asphalt due to drain down was in the range of 10-15 grams. It was observed that after addition of asphalt these mixes, especially FC-5 with Granite, they became very sensitive to the compaction temperature. For the mixes compacted without addition of asphalt, number of gyrations was determined as 50. However, for the saturated samples, the gyration level varied from 60-90. This indicates that the rate of reduction of temperature in case of polymer-modified asphalt is very high. Finally, after compaction the sample was not immediately retrieved from the mould. It was allowed to cool down for around 1 hr 15 min before ejecting the sample out (for unsaturated field mix the pills were allowed to cool down for 45 minutes before retrieving them). This was because it was observed that, because of high air voids, these mixes collapsed when they were retrieved from the mould immediately after compaction

PAGE 60

47 The final air void level and Gmm for the mixtures after compaction was as follows Table 4.7 Weight of PG 76-22 to be added Type Initial Final Weight of Mix (gms) Gmb Gmm Air Voids (%) Air Voids (%) %Gmm Gmm Asphalt Added (gms) Compaction Height (mm) FC-5 Limestone 4800 1.923 2.336 17.68 13.26 86.740 2.269 113.50 141.250 FC-5 Granite 4700 1.916 2.441 21.51 16.13 83.869 2.350 135.59 138.813 Novachip 4800 2.089 2.474 15.56 7.78 92.219 2.352 183.79 130.000 4.4 Long Term Oven Aging of Friction Course The mixes were taken from the wood way and hence they had already undergone short-term aging. The mixes were subjected to long-term aging according to LTOA-AASHTO PP2-94. However, the mixes being very course and open, there was a possibility of these mixes falling apart during aging. Hence, a procedure was developed to contain the compacted pills from falling apart during aging. A wired mesh with opening of 1/8 and steel clamp was used. The mesh size was chosen in order to ensure that there is good circulation of air within the sample for oxidation and at the same time, to prevent the smaller aggregate particles from falling off from the mesh. The specimen was rolled over twice by the mesh and two clamps were used to contain the specimen without applying excessive pressure on it. The whole system looked as shown in Figures 4.8-4.9

PAGE 61

48 Fig. 4.8 1/8 Mesh used for containing the Pill Fig. 4.9 Pill contained with Mesh

PAGE 62

CHAPTER 5 INDIRECT TENSILE TESTING The fracture evaluation of the friction courses was done within the framework of HMA fracture mechanics (16). The HMA fracture mechanics deals with the concept of threshold, which is defined as materials state between micro-damage and macro-crack development and is dealt in terms of energy. There are two energy limits, which define failure, viz., Dissipated Creep Strain Energy (DSCE) and Fracture Energy (FE). DSCE is chosen as criterion under repeated loading condition while FE is selected under critical loading condition as shown in Figure 4.1 Energy CASE 1 Repeated Load Cyclic Fatigue CASE 2 Fail at the Critical Load CASE 3 N (Number of Load Replications) Fail Fail N o Failure DEthreshold FEthreshold Fig. 5.1 Effect of Loading Condition 49

PAGE 63

50 The key points of the HMA Fracture Mechanics Model can be summarized as follows: If the threshold as defined in this model is not exceeded, micro damage is healable, once the threshold is exceeded, macro damage s not healable Under repeated loading conditions, DSCE can be used as threshold and it can be easily obtained from strength tests using the Superpave Indirect tensile test System (IDT) Asphalt being a viscoelastic material, crack imitation and propagation cannot be distinguished, cracks grow discontinuously (i.e. crack grows in step wise manner) Under critical loading condition, FE obtained from strength test can be used as threshold This model handles realistic loading condition and healing effects on asphalt pavements using the DSCE and FE as criterion All parameters needed to describe crack growth are obtained from relatively simple Indirect Tensile test (IDT) (i.e. resilient modulus, creep responsem-value, fracture energy to failure and tensile strength) 5.1 Sample Preparation The friction course being very porous, in order to avoid end effects, it was decided that the sample thickness be around 1.5 in. However, in case of FC-5 with limestone with additional asphalt, the thickness was reduced to 1.0 in because it was observed that due to high asphalt content, the limestone stone was being crushed at top before failure. Therefore, the thickness was reduced which then solved this problem. A cutting device, which has a cutting saw and a special attachment to hold the pills (Figure 5.2), was used to slice the pill into specimens of desired thickness. Two two-inch samples were from each specimen. Because the saw uses water to keep the blade wet, the specimens were dried for one day at room temperature to achieve the natural moisture content. Before testing, Specimens were placed in the humidity chamber for at least two days to negate moisture effects in testing. Gage points were attached to the samples using a steel template and vacuum pump setup and a strong adhesive (Figure 5.3). Four gage points were placed on each

PAGE 64

51 side of the specimens at distance of 19 mm (0.75 in.) from the center, along the vertical and horizontal axes. A steel plate that fits over the attached gage points was used to mark the loading axis with a marker. This helped placing the sample in the testing chamber assuring proper loading of the specimen (Figure 5.4). Fig. 5.2 Cutting Device

PAGE 65

52 Fig. 5.3 Gauge Point Attachment

PAGE 66

53 Fig. 5.4 Marking Loading Axes 5.2 Test Procedures Standard Superpave IDT tests (13,14) were performed on all mixtures to determine resilient modulus, creep compliance, m-value, D1, tensile strength, failure strain, fracture energy, and dissipated creep strain energy to failure. The tests were performed at 10C. 5.2.1 Resilient Modulus Test The resilient modulus is defined as the ratio of the applied stress to the recoverable strain when repeated loads are applied. The test was conducted according to the system developed by Roque et al (13) to determine the resilient modulus and the Poissons ratio. The resilient modulus test was performed in load control mode by applying a repeated haversine waveform load to the specimen for a 0.1 second followed by a rest period of 0.9 seconds. The load was selected to keep the horizontal strain in the linear viscoelastic

PAGE 67

54 range, in which horizontal strain is typically 150 to 350 micro-strains. The procedures for resilient modulus test are as follows: 1. The specimens compacted are cut parallel to the top and bottom faces using a water-cooled masonry saw to produce 2 inches thick specimens having smooth and parallel faces. 2. Four aluminum gage points are affixed with epoxy to each trimmed smooth face of the specimen. 3. Test samples are stored in a humidity chamber at a constant relative humidity of 60 percent for at least 2 days. In addition, specimens are cooled at the test temperature for at least 3 hours before testing. 4. Strain gauges are mounted and centered on the specimen to the gage points for the measurement of the horizontal and vertical deformations. 5. A constant pre-loading of approximately 10 pounds is applied to the test specimens to ensure proper contact with the loading heads before test loads are applied. The specimen is then tested by applying a repeated haversine waveform load for five seconds to obtain horizontal strain between 150 to 300 micro-strains. If the horizontal strains are higher than 050 micro-strains, the load is immediately removed form the specimen, and specimen is allowed to recover for a minimum 3minutes before reloading at different loading level. 6. When the applied load is determined, data acquisition program begins recording test data. Data are acquired at a rate of 150 points per seconds. 7. The resilient modulus and Poissons ratio are calculated by the following equations, which were developed based on three dimensional finite element analysis by Roque and Buttlar (1992). The equation is involved in the Superpave Indirect Tensile Test at Low Temperatures (ITLT) program, which was developed by Roque et al (1997). MR= H t D C co m p P GL Where, M R = Resilient modulus P = Maximum load GL = Gauge Length H = Horizontal Deformation t, D = Thickness, Diameter C comp = 0.6354 (X/Y) -1 -0.332

PAGE 68

55 5.2.2 Creep Test Creep compliance is a function of time-dependent strain over stress. The creep compliance curve was originally developed to predict thermally induced stress in asphalt pavement. However, because it represents the time-dependent behavior of asphalt mixture, it can be used to evaluate the rate of damage accumulation of asphalt mixture. As shown in Figure 5.5, D0, D1, and m-value are mixture parameters obtained from creep compliance tests. Although D1 and m-value are related to each other, D1 is more related to the initial portion of the creep compliance curve, while m-value is more related to the longer-term portion of the creep compliance curve. The m-value has known to be related to the rate of damage accumulation and the fracture resistance of asphalt mixtures. In other words, the lower the m-value, the lower the rate of damage accumulation. However, mixtures with higher m-value typically have higher DCSE limits. The creep compliance is a time dependant strain, (t), divided by a constant stress. That is, the inverse of the creep compliance, which is called creep stiffness, is a kind of stiffness. According to the analysis conducted by Roque et al (13), MR is higher than creep compliance stiffness at 1 second. The Superpave Indirect Tensile Test at Low Temperatures (ITLT) computer program was used to determine creep properties of the mixtures. The test was conducted in a load control mode by applying a static load. The load was selected to keep the horizontal strain in the linear viscoelastic range, which is below a horizontal strain of 500 micro strains. The test procedure was presented by Roque et al (1997). The procedures for indirect tensile creep test consist of the following steps:

PAGE 69

56 The preparation of test samples and the pre-loading are same as those for resilient modulus test Apply a static load for 1000 seconds. If the horizontal deformation is greater than 180 micro inch at 100 seconds, the load is immediately removed from the specimen, and specimen is allowed to recover for a minimum 3 minutes before reloading at a different level. At 100 sec, the horizontal deformation should be less than 750 micro inches When the applied load is determined, the data acquisition program records the loads and deflections at a rate of 10 Hz for the first 10 seconds, 1Hz for the next 290 seconds, and 0.2 Hz for the remaining 700 seconds of the creep test. The computer program, ITLT, was used to analyze the load and deflection data to calculate the creep compliance properties. Creep compliance and Poissons ratio are computed by the following equations. P GL D( t ) = H t D Ccom p = -0.1+1.480 (X/Y) 2 0.778 (t/D) 2 (X/Y) 2 Where, D (t) = Creep Compliance 5.2.3 Strength test Failure limits such as tensile strength, failure strain, and fracture energy were determined from strength tests using the Superpave IDT. These properties are used for estimating the cracking resistance of the asphalt mixtures. The strength test was conducted in a displacement control mode by applying a constant rate of displacement of 50 mm/min for field mix and 100 mm/min for saturated mix until the specimens failed. The horizontal and vertical deformation and the applied load are recorded at the rate of 20 Hz during the test. The maximum tensile strength is calculated as the following equation. b d 2 P C sx S t =

PAGE 70

57 Fig. 5.5 Power Model for Creep Compliance

PAGE 71

58 Where, S t = Maximum Indirect tensile Strength P = Failure load at first crack C sx = 0.948 0.01114 (b/D) 0.2693 + 1.436 (b/D) b, D = Thickness, diameter From the strength test and the resilient modulus test, fracture energy and dissipated creep strain energy can be determined. Fracture energy is a total energy applied to the specimen until the specimen fractures. Dissipated creep strain energy (DCSE) is the absorbed energy that damages the specimen, and dissipated creep strain energy to failure is the absorbed energy to fracture (DCSEf). As shown in the Figure 3.9, fracture energy and DCSEf can be determined as described below. The ITLT program also calculates fracture energy automatically. MR f-St M R = => 0 = S t f 0 MR Elastic Energy (EE) = () S t ( f 0 ) Fracture energy (FE) = S () d f 0 Dissipated Creep Strain energy (DSCE) = FE EE Where, S t = Tensile Strength f = Failure Tensile Strain

PAGE 72

59 Fig.5.6 FE and DCSE from Strength Test 5.3 Issues Related with IDT Testing of Friction Course As already mentioned, the specimen thickness were kept between 1.5 2 in for the all field mixes and for saturated FC-5 Granite and Novachip to avoid end effects. The reason, this thickness range was chosen was because these graded friction courses are so open that at the point of contact with the loading strip, the surface might be so weak that local failure might occur under stress concentration. Thus, there was a possibility of not getting true tensile failure. However, for saturated FC-5 limestone, the thickness was kept close to one inch because with 2-inch sample the aggregate at the top was being crushed. This was because the mix had such high asphalt content that, at 10C, the polymer-modified asphalt became stiff and before the failure could propagate through the mastic, the lime rock at the top was crushed. Hence, the thickness was reduced to lower the peak

PAGE 73

60 load to failure and avoid end effects. Another reason for the end effect failure in limestone is the loading rate, which has been discussed in the following paragraph. Now for the field mix, the loading rate in strength test was 50-mm/ min. However, for saturated sample, the samples were failing due to end effects at that loading rate. This was observed primarily because of shear failure rather than true tensile failure. Hence, the loading rate was increased to 100 mm/min, and end effect failure was no longer observed.

PAGE 74

CHAPTER 6 MIX DESIGN APPROACH FOR GEORGIA PEM The Georgia DOT uses Marshalls blow for design of PEM. However, for the Georgia PEM mix design, we used Superpave Gyratory Compactor. Further, the surface capacity determination, as explained in literature review, is not needed if the gradation is within the desired gradation band. The over all mix design approach is explained in the following chapter. 6.1 Evaluation of compaction Level for Georgia PEM The compaction level used for Georgia PEM was 50 gyrations. Now, as already stated, the approach used in determination of compaction level was based on rate of change of the slope of compaction curve. The Locking Point was identified as the point at which two gyrations at same gradient of slope were preceded by two gyrations at same gradient of slope. Thus based on above definition, the locking point for the three pills of Georgia PEM was identified as given in Table 6.1 Thus, from the table its clear that the compaction level for Georgia PEM is also around 50 gyrations, which was used for the design. 6.2 Mix Design Approach Scope The design for the Georgia PEM consists of three steps. The first step is to conduct AASHTO T-245 to determine asphalt cement content then, secondly to determine optimum asphalt content. The final step is performing GDT-127 61

PAGE 75

62 Table 6.1 Locking Point for Georgia PEM Sample No. of Gyrations Gradient of the Slope 1 43 0.0015 44 0.0015 45 0.0014 46 0.0014 2 44 0.0014 45 0.0014 46 0.0013 47 0.0013 3 47 0.0011 48 0.0011 49 0.0010 50 0.0010 Initial Asphalt Content Based on number of experiments, the GDOT suggests using 6 % as the first estimate of asphalt content. It has been observed that if the gradation is within the specified limits, the initial estimate is comes out to be 6%. Modified Marshall Design and Optimum AC : 1. Heat the coarse aggregate, the mould to and the AC to 330 F 3.5 F 2. Mix aggregate with asphalt at three asphalt contents, viz., 5.5%, 6% and 6.5%. Just before mixing, add the required amount of mineral fibers to the aggregate. Prepare three samples at each of the asphalt content 3. After mixing, return to oven for two hours for STOA at 320 F 3.5 F Then compact using the Superpave Gyratory Compactor 50 gyrations 4. When compacted, cool to the room temperature before removing from the mold. It typically takes 1 hour 45 min to cool down. 5. Bulk Specific Gravity: 6. Determine the density of a regular shaped specimen of compacted mix from its dry mass (in grams) and its volume in cubic centimeters obtained from its dimensions for height and radius. Convert the density to the bulk specific gravity by dividing by 0.99707 g/cc, the density of water at 25 C 7. Bulk Sp.Gr = W / ( r2h/ 0.99707) = Weight (gms) 0.0048417/Height (in) W = Weight of specimen in grams 8. R = radius in cm 9. H = height in cm 10. Calculate percent air voids, VMA and voids filled with asphalt based on aggregate specific gravity

PAGE 76

63 11. Plot VMA curve versus AC content 12. Select the optimum asphalt content at the lowest point on VMA curve Drain-Down Test Perform the drain test in accordance w ith the GDT 127 (Method for determining Drain Down characteristics in Uncompacted Bituminous Mixtures). A mix with an optimum AC content as calculated above is pl aced in a wired basket having 6.4 mm (1/4 inch) mesh openings and heated 14 C (25 F) above the normal production temperature (typically around 350F) for one hour. The amount of cement, which drains from the basket, is measured. If the sample fails to meet the requirements of maximum draindown of 0.3 %, increase the fiber cont ent by 0.1 % and repeat the test. Thus based on the above design procedur e, the optimum asphalt content was determined to be 6%. The fiber content was 0.4% of the total mix. The drain down was minimal (around .01%). The final mix design in as shown in Fig. 6.1 It is recommended by GDOT that the as phalt content should not be below 6% because of coating issues. The film thickne ss at 6% asphalt was around 39.6 micron. The drain down was .01%, which is well below th e limit. Thus based on 6% AC, pills were prepared for IDT testing.

PAGE 77

64 Effective Sp Gravof Agg% ACGmmGmbVMAVTMVFA2.6415.52.4421.93630.7420.7232.6062.4141.96130.2318.7837.866.52.3891.96730.3817.6841.82Optimum AC6% Mixing Temperature330 F Mineral Fiber0.4% of Total Mix Compaction Temperature325 F VMA30.1030.2030.3030.4030.5030.6030.7030.805.45.65.866.26.46.6% ACVMA VTM17.0018.0019.0020.0021.005.45.65.866.26.46.6% ACVTM VFA30.0032.0034.0036.0038.0040.0042.0044.0055.566.57% ACVMA Fig 6.1 Final Mix Design for Georgia PEM

PAGE 78

CHAPTER 7 FINDINGS AND ANALYSIS The results obtained from the IDT test have been analyzed using ITLT software developed at University of Florida. Using the results from the software DCSE and Energy Ratio were calculated. Energy Ratio is dimensionless parameter that serves as single criteria for cracking performance of mixtures in pavements. It is defined as follows: ER = DCSE f / DCSE min Where, DCSE f is the dissipated creep strain energy threshold, and DCSE min is the minimum dissipated creep strain energy, which is given by: DCSE min = m 2.98 D 1 / A, Where, m and D 1 are creep compliance parameters. A is a parameter dependent on tensile stress and tensile strength, St A = 0.0299 -3.10 (6.36-St) + 2.46 10 -8 Where, is the applied tensile stress. The applied tensile stress is taken at the bottom of the AC layer and is very much dependent on the stiffness of the AC layer. In case of friction course, the whole pavement structure can be regarded as a composite consisting of friction course and AC layer beneath it. Thus, the stress at the bottom of the AC layer should be considered for ER calculation for the friction course. It should be noted that the ER was determined for dense graded mixtures only and it has not been truly calibrated for friction courses. Thus, a typical pavement structure with the given loading condition was considered for determining ER for all the mixes. This has been illustrated in Fig7.1. 65

PAGE 79

66 Friction Course Mr ( from IDT ) AC Layer, E = 800000 psi = 0.4 Base course E = 500000 = 0.35 Subgrade, E = 10000 psi Pressure = 120 psi, Load = 9000 lbs 0.75 in 8 in 12 in Fig 7.1 Typical Pavement Structure for Stress Calculation The thickness for FC-5 and Novachip was taken as 0.75 in and for Georgia PEM it was taken as 1.25 in. The stresses were calculated using KENLAYER and the results have been attached in Appendix C. The other parameters used for evaluation are the FE, DCSE, tensile strength, failure strain and creep strain rate. The creep strain rate for a 1000 sec creep test is calculated as follows: d(t)/dt/ = dD(t)/dt = D 1 m (1000) m-1 Where, d(t)/dt/ = strain rate per unit stress D(t) = creep compliance 7.1 Evaluation of FC-5 and Novachip Field Mix The test results for the friction courses are presented in Figures 7.1 through 7.7. The three field mixes were also subjected to long-term oven aging to study the effects of aging on their fracture resistance. In addition, to replicate the interface between the

PAGE 80

67 surface mixture and the underlying HMA, which contains the polymer, modified tack coat, additional PG76-22 was added to the field mix to reach the desired target air void level. The results for the saturated mixes are presented in Figures 7.15 through 7.21 7.1.1 Unaged Field Mix As shown in Fig 7.4, for FC-5 Granite and FC-5 Limestone, the ER is around 1.5 which indicates good field performance. For Novachip, the ER is close to 3, which indicates that it is a better performing surface course than FC-5. In case of FC-5 Granite, the failure strain is around 3000 micro strain and DCSE is close to 3 KJ/m3. For limestone, the DCSE is very low (just below 1 KJ/m3). This is primarily because of low failure strain (around 950 micro strain). For Novachip the DCSE is around 4.5 KJ/m3 with a high failure strain of 3200 micro strain. As far as the tensile strength and resilient modulus are concerned, Novachip has higher values as compared to FC-5 granite and limestone. Finally, Novachip and FC-5 Granite have significantly high creep rate (close to 3e-08/psi-sec) as compared to FC-5 Limestone. For Granite, this is one of the reasons for low ER value. For Novachip high creep rate is compensated by high DCSE, which in turn is reflected in its high ER value. 7.1.2 Aged Field Mix In case of granite, there is substantial decrease in ER due to long-term aging. As shown in Fig 7.5, for granite-aged mix, the failure strain reduced by half as compared to unaged mix. As a result, the DCSE also plummeted to 1 KJ/m 3 which resulted in really low ER value of 0.7.

PAGE 81

68 0.000.501.001.502.002.503.003.504.004.505.00FC-5 GFC-5 LNOVACHIPFE (KJ/m3) Unaged Aged Fig 7.2 Fracture Energy for Field Mix 0.000.501.001.502.002.503.003.504.004.505.00FC-5 GFC-5 LNOVACHIPDCSE (KJ/m3) Unaged Aged Fig 7.3 Dissipated Creep Strain Energy for Field Mix

PAGE 82

69 0.000.501.001.502.002.503.003.504.004.50FC-5 GFC-5 LNOVACHIPEnergy Ratio Unaged Aged Fig 7.4 Energy Ratios for Field Mix 0.0500.01000.01500.02000.02500.03000.03500.04000.0FC-5 GFC-5 LNOVACHIPef (micro strain) Unaged Aged Fig 7.5 Failure Strain for Field Mix

PAGE 83

70 0.000.501.001.502.002.50FC-5 GFC-5 LNOVACHIPTensile Strength (MPa) Unaged Aged Fig 7.6 Tensile Strength for Field Mix 0.001.002.003.004.005.006.007.008.009.0010.00FC-5 GFC-5 LNOVACHIPMr (GPa) Unaged Aged Fig 7.7 Resilient Modulus for Field Mix

PAGE 84

71 0.00E+005.00E-091.00E-081.50E-082.00E-082.50E-083.00E-083.50E-084.00E-084.50E-085.00E-08FC-5 GFC-5 LNOVACHIPStrain Rate/stress(1/psi-sec) Unaged Aged Fig 7.8 Creep Rate for Field Mix The creep rate however increases with aging for granite. This could be attributed to the fact there are critical points in the matrix of aggregate and mastic, wherein aging takes place locally. These critical points are points of stone-stone contact. Further, there are regions in the matrix where aging is not so effective due to large globule of mastic. Hence, due to aging, we have a structure with varied stiffness and hence different regions would respond differently to load application. Thus, any kind of damage to these critical points, which govern the creep behavior of the specimens, causes damage in the specimens, resulting in higher creep rate. This is illustrated in Fig 7.9 (a). For Novachip, aging seems to have little effect on FE, DCSE and ER. However, interestingly, for limestone, aging increases ER by factor of two (ER for aged limestone is around 4.2). This is primarily due to higher failure strain, DCSE, and lower creep rate. To explain this phenomenon, a study was done on the effect of aging on these mixes, which is presented below.

PAGE 85

72 The asphalt content of the friction course is around 6 %, which is high as compared to any dense graded mixture. Further, limestone in general has a very rough texture and has lot of crevices and pores in it. Now, the long-term oven aging temperature is around 85 C and at such a high temperature, the asphalt flows, occupies the crevices and the pores in the aggregate. This is illustrated in Fig 7.9 (b). Thus, essentially, with aging in laboratory, the absorption by limestone increases and the aggregates become more ductile. In addition, the bonding between the aggregates also increases. Now, visual observation of the cracked specimens indicated that the failure happened through the aggregate. Thus, because of increased absorption due to aging, limestone becomes more ductile and hence the failure strain increases resulting in increased DCSE and ER. It should however be noted that in the field, the temperature varies from anywhere between 10-60 C and only in summers the pavement temperature approaches the higher end. In addition, these mixtures being surface mixtures will certainly not be subjected to very high temperatures. Thus, we believe that it is only in laboratory aging, such results can be observed and that it is not a realistic representation of field condition To corroborate the laboratory results further, another set of testing was done on limestone under the following conditions i.e. 1) Unaged, 2) 2-Day aging, 3) 5-Day aging (LTOA) 4) 10-Day aging. The results are presented in Figures 7.10 through 7.16. The results for unaged and long-term oven aged specimens showed the same trend as before. Clearly, looking at the ER, we see that 2-Day and 5-Day aged mix have higher ER as compared to the unaged mix. The same is the case with DCSE and failure strain. However, for the 10-day aged mix, the DCSE falls below that of unaged mix, though the ER remains higher.

PAGE 86

73 Area where aging might not be effective Mastic Critical Point where maximum aging will occu r (A) Absorbed asphalt in Crevices (B) Fig 7.9 Effect of aging on A) FC-5 Granite B) FC-5 Limestone

PAGE 87

74 The 2-day aged mix has the highest ER and then it decreases with aging. The same trend is observed for DCSE and failure strain. Thus, as already stated before, during the process of laboratory aging, the asphalt starts flowing and occupies the crevices and pores in the lime rock making the mix more ductile. However, as aging progresses, the absorbed asphalt also starts getting aged and we see decrease in the failure strain, and hence the DCSE, for the 5-day and 10-day aged mix. This again is reflected in the lower ER values. The resilient modulus increases slightly with aging. The tensile strength increases for the aged mixes. For this new set of testing, the creep rate of the unaged mix seems to be on the higher side as compared to field mix. However, the 5-day aged mix seems to have around the same creep rate as obtained before for aged field mix. The 10day aged mix has the lowest creep rate. 7.1.3 Saturated Unaged Mix As already mentioned, modified PG76-22 was added to the mixes to see the effect of Novabond on the fracture resistance of the mixes. Clearly, for FC-5 granite and Novachip, Novabond increases the FE, DCSE and failure strain (Fig 7.18-7.24). Infact, for Novachip, the FE increases to 9 KJ/m 3 which is twice as much as for unaged field mix. The ER for granite (around 2) does not increase substantially, possibly because of increase in creep rate with the addition of asphalt (Fig 7.17). However, for Novachip, the creep rate remains more or less same as that of the field mix and DCSE increases significantly to 9 KJ/m 3 hence we see a huge jump in the ER value (slightly above 6). For limestone, the failure strain increases to around 1400 micro strain and DCSE increases slightly to around 1.2 KJ/m3. The ER for limestone saturated however decreases as compared to the field mix. This is primarily because of the high creep rate (around 3e

PAGE 88

75 08/psi-sec) due to addition of asphalt. It should be noted that for the saturated mixes, the resilient modulus decreases due to increased asphalt content. 7.1.4 Saturated Aged Mix As in case of field mix, aging decreases ER for Granite (slightly above 1) and Novachip. This is because of significant decrease in the failure strain and hence in the DCSE. In case of limestone, the aged mix has higher ER than the unaged, as was the case with field mix. However, the aged limestone mix has lower DCSE and failure strain than unaged mix. The reason for higher ER is because of high creep rate of unaged saturated mix. The creep rate of aged saturated limestone mix was same as the field mix. Another interesting thing to note is that, the FE, DCSE and ER of aged field limestone mix are much greater than saturated limestone mix. A possible explanation for this is that, at high asphalt content, the limestone loses the benefit of its rough surface texture and the mix with aging becomes very brittle because of high asphalt content. This results in lower failure strains and hence lower fracture energy. 0.000.200.400.600.801.001.201.401.601.802.00Unaged2 Day Aging5 Day Aging10 Day AgingFE (KJ/m3) Fig. 7.10 FE for FC-5 Lime at Various Stages of Aging

PAGE 89

76 0.000.200.400.600.801.001.201.401.601.802.00Unaged2 Day Aging5 Day Aging10 Day AgingDCSE (KJ/m3) Fig. 7.11 DCSE for FC-5 Lime at Various Stages of Aging 0.0200.0400.0600.0800.01000.01200.01400.01600.0Unaged2 Day Aging5 Day Aging10 Day Agingef Fig. 7.12 Failure Strain for FC-5 Lime at Various Stages of Aging

PAGE 90

77 0.000.501.001.502.002.503.003.504.00Unaged2 Day Aging5 Day Aging10 Day AgingEnergy Ratio Fig. 7.13 Energy Ratio for FC-5 Lime at Various Stages of Aging 0.000.200.400.600.801.001.201.401.601.802.00Unaged2 Day Aging5 Day Aging10 Day AgingTensile Strength (MPa) Fig. 7.14 Tensile Strength for FC-5 Lime at Various Stages of Aging

PAGE 91

78 0.001.002.003.004.005.006.007.008.009.0010.00Unaged2 Day Aging5 Day Aging10 Day AgingMr (GPa) Fig. 7.15 Resilient Modulus for FC-5 Lime at Various Stages of Aging 0.00E+002.00E-094.00E-096.00E-098.00E-091.00E-081.20E-081.40E-08Unaged2 Day Aging5 Day Aging10 Day AgingCreep Rate/Stress(1/psi) Fig. 7.16 Creep Rate for FC-5 Lime at Various Stages of Aging

PAGE 92

79 0.001.002.003.004.005.006.007.00FC-5 GFC-5 LNOVACHIPEnergy Ratio SaturatedUnaged Field MixUnaged Fig. 7.17 Comparison of ER between unaged Saturated and Field mix 0.001.002.003.004.005.006.007.008.009.0010.00FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)FE (KJ/m3) Unaged Aged Fig. 7.18 FE for Saturated Mix

PAGE 93

80 0.001.002.003.004.005.006.007.008.009.0010.00FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)DCSE (KJ/m3) Unaged Aged Fig. 7.19 DCSE for Saturated Mix Failure Strain(Saturated)0.01000.02000.03000.04000.05000.06000.0FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)ef Unaged Aged Fig. 7.20 Failure Strain for Saturated Mix

PAGE 94

81 0.001.002.003.004.005.006.007.00FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)Energy Ratio Unaged Aged Fig 7.21 ER for Saturated Mix 0.000.501.001.502.002.50FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)Tensile Strength (MPa) Unaged Aged Fig. 7.22 Tensile Strength for Saturated Mix

PAGE 95

82 012345678910FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)Mr (GPa) Unaged Aged Fig. 7.23 Resilient Modulus for Saturated Mix 0.00E+001.00E-082.00E-083.00E-084.00E-085.00E-086.00E-087.00E-088.00E-08FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)Strain Rate/stress(1/psi-sec) Unaged Aged Fig. 7.24 Creep Strain Rate for Saturated Mix

PAGE 96

83 7.1.5 Moisture Conditioning The FC-5 mixes were subjected to moisture conditioning as outlined in AASHTO-T283. The procedure is follows: For each of the mixes, the pills were compacted to 50 gyrations. The pills were then placed in a vacuum bath to remove the entrapped air. Then, they were conditioned by placing them in a water bath at 60C for 24 hours. After this, the samples were again conditioned in a 25C water bath for two hours. Then water was allowed to drain down for 36 hrs before cutting the samples. The samples were then tested and the results are as presented below: In case of FC-5 Granite, two of three samples seemed to have failed in the creep test itself. The creep test results for the samples are presented in Fig 7.25 1 As evident from the figures, for a load as low as 4 pounds, the samples were having a deformation close to 900 micro inches. Infact at that low a load, there was no elastic response and the creep curve did not seem to follow the power model. Hence, the creep rate for all the samples was calculated using the straight-line portion of the creep deformation curve and the results were compared to that of the unconditioned samples (Fig 7.25) Clearly from Fig 7.26 it is evident that the creep rate for conditioned samples increased significantly and that the specimens failed during creep test itself. Hence, the strength test results are not truly representative of the behavior of the mix. Further, the resilient modulus of the conditioned samples was lower as compared to unconditioned samples (close to 3.6 M Pa). This clearly indicates that the samples had been damaged during conditioning. Thus FC-5 Granite is highly sensitive to moisture conditioning. 1 Refer Appendix C for all the creep test results

PAGE 97

84 00.00010.00020.00030.00040.00050.00060.00070.00080.0009020040060080010001200t (Sec)Deformation (Micro Inches) x (A) 00.00020.00040.00060.00080.0010.0012020040060080010001200Time (Sec)Deformation (inches) x (B) Fig. 7.25 Creep Deformation Curve for Conditioned samples of FC-5 with Granite

PAGE 98

85 0.000E+005.000E-071.000E-061.500E-062.000E-062.500E-063.000E-063.500E-06G12G21G22Gcon11Gcon12Gcon22Creep Rate/Unit Stress Fig. 7.26 Creep Rate for Unconditioned and Conditioned samples FC-5 Granite Moisture conditioning helps aging the mix and hence we see an increase in Mr and tensile strength for conditioned FC-5 Limestone samples. Further, due to increased absorption during conditioning, we see an increase in FE and ER for conditioned samples. In addition, the creep rate decreases significantly for the conditioned samples. The results are presented in Figures 7.27 through 7.32. 00.20.40.60.811.2UnconditionedConditionedFE (Kj/m3) Fig.7.27 FE after Conditioning for FC-5 with Limestone

PAGE 99

86 0.000.100.200.300.400.500.600.700.800.901.00UnconditionedConditionedDCSE (Kj/m3) Fig.7.28 DCSE after Conditioning for FC-5 with Limestone 0.000.501.001.502.002.503.00UnconditionedConditionedER Fig.7.29 ER after Conditioning for FC-5 with Limestone

PAGE 100

87 020040060080010001200UnconditionedConditionedef (micro Strain) Fig.7.30 Failure Strain after Conditioning for FC-5 with Limestone 012345678910UnconditionedConditionedMr (MPa) Fig.7.31 Mr after Conditioning for FC-5 with Limestone

PAGE 101

88 0.00E+001.00E-092.00E-093.00E-094.00E-095.00E-096.00E-097.00E-09UnconditionedConditionedCreep Rate/Unit Stress Fig.7.32 Creep Rate after Conditioning for FC-5 with Limestone 7.2 Evaluation of Georgia PEM In case of Georgia PEM samples, it was observed that the samples had exceedingly high creep rate when tested immediately after Resilient Modulus Test. However, when there was a sufficient waiting period between the Mr and creep tests, the samples behaved normally. This indicates that the samples take lot of time to recover from delayed elasticity and that a waiting period of 30 minutes should be observed in between the tests. The test results are presented in Fig. 7.33 through 7.39. The FE and DCSE for unaged Georgia PEM are very high and comparable to Novachip (DCSE is close to 4 KJ/m3). They have exceedingly high failure strain of around 4500 micro strains, which is much higher than FC-5 and Novachip. This probably the effect of polymer modified asphalt in conjunction with the gradation. Interestingly, the ER is just around 2. This is because of a very high creep rate of 1e-07/psi-sec. One of the possible reasons for high creep rate may be attributed to segregation of fines from

PAGE 102

89 coarse aggregate during mixing. The segregation results in improper coating of fines on coarse aggregate. The final matrix of aggregate and asphalt essentially consists of globules of mastic and coarse aggregate rather than a uniform coating of mastic over coarse aggregate. Thus, the structure loses the inherent strength, which the fines provide to the mix. Hence, we see such a high creep rate (which indicates damage) in case of the laboratory prepared Georgia PEM mixes. This is illustrated in Fig 7.32. The PEM seems to be very sensitive to aging. The LTOA results show that the FE plummets down to 1 KJ/m3. The failure strain decreases to 1500 micro strains. The result is very similar to that of FC-5 Granite with both of them having very high air voids. This makes these mixes highly prone to aging. The resilient modulus more or less remains same as the unaged mix. There is a slight decrease in tensile strength and this could be the effect of aging. The ER decreases to 1 and creep rate comes down to 5e-08/psi-sec. Well-reinforced aggregate Well-distributed mastic with uniform coating over aggregate Loosely reinforced Aggregate Globules of Mastic (A) (B) Fig 7.32 Matrix of A) Loosely aggregates with globules of mastic B) Well reinforced aggregate uniform coating of mastic

PAGE 103

90 0.000.501.001.502.002.503.003.504.004.505.00FC-5 GFC-5 LNOVACHIPGPEMFE (KJ/m3) Unaged Aged Fig.7.33 FE for Georgia PEM 0.000.501.001.502.002.503.003.504.004.505.00FC-5 GFC-5 LNOVACHIPGPEMDCSE (KJ/m3) Unaged Aged Fig.7.34 DCSE for Georgia PEM

PAGE 104

91 0.0500.01000.01500.02000.02500.03000.03500.04000.04500.05000.0FC-5 GFC-5 LNOVACHIPGPEMef Unaged Aged Fig.7.35 Failure Strain for Georgia PEM 0.000.501.001.502.002.503.003.504.00FC-5 GFC-5 LNOVACHIPGPEMER Unaged Aged Fig.7.36 ER for Georgia PEM

PAGE 105

92 0.000.200.400.600.801.001.201.401.601.802.00FC-5 GFC-5 LNOVACHIPGPEMSt (MPa) Unaged Aged Fig.7.37 Tensile Strength for Georgia PEM 0.001.002.003.004.005.006.007.008.009.0010.00FC-5 GFC-5 LNOVACHIPGPEMMr (1/GPa) Unaged Aged Fig.7.38 Resilient Modulus for Georgia PEM

PAGE 106

93 0.00E+002.00E-084.00E-086.00E-088.00E-081.00E-071.20E-07FC-5 GFC-5 LNOVACHIPGPEMCreep Rate (1/psi-sec) Unaged Aged Fig.7.39 Creep Rate for Georgia PEM

PAGE 107

CHAPTER 8 SUMMARY, FINDINGS AND ANALYSIS 8.1 Summary of Findings Three different mixes from the field, viz., FC-5 with Granite, FC-5 with Limestone and Novachip, were tested for evaluation of their fracture resistance. It was observed that this friction exhibit good level of interlocking at low compaction level and that they reach the locking point at low gyrations. Evaluation of these mixes showed that Novachip was the best performing mix among the three mixes. Further, laboratory aging seems to be an issue with limestone. Finally, Georgia PEM mixes were designed and tested for fracture evaluation. The same mix design as used by Georgia DOT was used except for the introduction of Superpave Gyratory compactor for compaction instead of traditional Marshall blows. The unaged Georgia PEM mix had FE and failure strain comparable to Novachip. The summary of the findings is as follows: Locking point as defined by Vavrik did not seem to be valid for friction course since it was identified at very high number of gyrations. For friction course, the rate of change of compaction served as a better indication of resistance to compaction and the compaction level predicted using this concept seemed reasonable. In case of mixes with additional PG 76-22, the mixing temperature was increased to 320F and compaction temperature was reduced to 255F to avoid draindown. The additional asphalt added was such that the air voids were reduced by 50% for Novachip and 25 % for FC-5 limestone and Granite. For the IDT tests, the specimen thickness was kept around 1.5-2 inches in order to avoid end effects during failure. In case of saturated sample, the loading rate was increased to avoid shearing of the samples and to obtain a true tensile failure. The fracture analysis on the field mixes, showed that Novachip was superior to FC-5 limestone and Granite. In case of limestone, laboratory aging seems to increase its fracture resistance. This probably because of increased absorption during aging, as a result enhancing the fracture properties. FC-5 Granite seems to be highly sensitive to aging. 94

PAGE 108

95 The addition of the polymer-modified binder, PG 76-22, seems to increase the fracture resistance in case of Novachip. For FC-5 granite, even though the FE and DCSE increases because of significant increase in creep rate, the ER value remains more or less same as that of the unsaturated mix. In case of FC-5 Granite, the moisture conditioned samples failed in the creep test itself. For FC-5 limestone, conditioning seems to improve the fracture resistance. This is again because of increased absorption. For Georgia PEM, 50 gyrations were used for compaction instead of 50 Marshall blows, which the GDOT uses for PEM. It was observed that, the PEM samples took lot of time to recover from delayed elasticity. Hence, a waiting period of 30 minutes was used between Mr, creep and strength test. Unaged Georgia mix had very high FE and failure strain which were comparable to Novachip. The ER though was just around same as FC-5 because of the high creep rate. In addition, the Georgia PEM mix was highly sensitive to aging and had very low FE and failure strain. The ER for aged mix was just below one. 8.2 Conclusions Rate of change of compaction serves is a better way for identifying locking point in case of friction courses. Based on this study, compaction level for friction courses was identified as 50 gyrations. For saturated samples, higher loading rate is desired to obtain a true tensile failure. In general, in case of friction course, there are certain critical points in the matrix, which have maximum exposure to aging and there are certain areas where aging is not that effective. These critical points dictate the fracture resistance of the mix as is observed in case of FC-5 Granite and Georgia PEM. In case of FC-5 with limestone, laboratory aging helps absorption, which in turn increases the fracture energy and the ER. Hence, laboratory aging is not a realistic representation of field aging since in field; aging does not really enhance the absorption capacity of limestone. Addition of PG 76-22 increases ER significantly for Novachip. However, for FC-5 Granite and Limestone, there is not much of an improvement in ER value. In case of FC-5 Limestone and Granite, the ER values may not truly represent the effect of novabond. The Novabond is supposed to be effective in stress relief, provides better bonding and resists aging of the underlying layer. FC-5 Granite is highly moisture sensitive whereas moisture conditioning improves the fracture resistance of FC-5 Limestone. Unaged Georgia PEM has extremely high fracture energy and failure strain which is comparable to Novachip. However, the ER was not very high due to its high creep rate. It is also highly sensitive to aging.

PAGE 109

96 8.3 Recommendations Novabond may be used for friction courses to increase the fracture resistance For FC-5 with limestone, the absorption should be enhanced which will result in improved fracture resistance. Thus, use of softer binder is recommended. In case of aged FC-5 limestone, microscopic images of the failed specimens should be looked at to validate the hypothesis of absorption in limestone. ER was traditionally developed for dense graded mixes. Hence, further study must be done to calibrate ER for friction course. Mixtures with high fracture energy but not so high ER may need to be evaluated further. The mixing procedure in the laboratory may be refined for the friction courses to avoid segregation of fines. For calculating film thickness in case of Georgia PEM mixes with Limestone, absorption should be taken into account. Hence, the correction to the film thickness must be done accordingly.

PAGE 110

APPENDIX A VOLUMETRICS FOR FC-5 AND NOVACHIP

PAGE 111

Table A.1 Gradation for FC-5 Granite, Limestone and Novachip 98 Type Granite Limestone Novachip Sieve Size (mm) #7 #789 Granite JMF Screens S1A S1B Screens JMF #7 #789 Granite JMF Screens % by weight 77 12 11 100 60 10 30 100 45 48 7 100 19 100 100 100 100 100 100 100 100 100 100 100 100 12.5 95 100 100 96 95 100 100 91 79 100 100 97 9.5 64 92 100 75 62 93 100 67 36 92 100 77 4.75 11 20 97 22 11 20 95 23 7 26 100 37 2.36 3 5 68 10 6 8 72 10 3 7 68 26 1.18 2 3 43 7 4 5 50 7 3 3 67 18 0.6 2 3 28 5 3 4 33 7 3 3 55 12 0.3 2 3 18 5 2 3 21 5 3 2 35 8 0.15 2 3 11 4 2 2 14 5 2 2 14 6 0.075 1.1 2.5 8 3.1 2 2 8.5 4 1 1 3 4

PAGE 112

Table A.2 Rice Test for the Friction Courses 99 Sample Flask Number A Weight of flask A Wt of flask, & Water A Wt. Of Sample A Wt of Sample in Water A Multiplier A Wt of Sample in Water Corr. A Wt. Dry Back (SSD) Combined Of Sample in Water Corr. Lab GMM Average Gmm 6 1777.8 6079.2 1501.5 855.2 1.001259 856.27 1501.5 856.27 2.327 2.324 FC-5 Lime 7 1889.1 6121.87 1504.3 861.73 1.001259 862.81 1504.6 862.81 2.344 6 1777.8 6079.2 1513.5 893.1 1.000705 893.72 1513.5 893.72 2.442 2.441 FC-5 Granite 7 1889.1 6121.87 1504 886.93 1.000636 887.49 1504 887.49 2.439 6 1777.8 6079.2 1505.3 895.4 1.001017 896.31 1504.7 896.31 2.473 2.474 Novachip 7 1889.1 6121.87 1509.9 898.93 1.001039 899.86 1509.9 899.86 2.475 FC-5 Lime 6 1777.8 6079.2 1498.4 852.2 1.000705 852.80 1498.8 852.80 2.320

PAGE 113

100 Type FieldMix SaturatedMix Weight of Mix (gms) Gmb Gmm Air Voids(%) Air Voids (%) %Gmm Gmm AsphaltAdded (gms) Compaction Height (mm) FC-5 Limestone 4800 1.923 2.336 17.68 13.26 86.740 2.269 113.50 141.250 FC-5 Granite 4700 1.916 2.441 21.51 16.13 83.869 2.350 135.59 138.813 Novachip 4800 2.089 2.474 15.56 7.78 92.219 2.352 183.79 130.000 Table A.3 Air voids and Asphalt content for Friction course

PAGE 114

APPENDIX B VOLUMETRICS FOR GEORGIA PEM

PAGE 115

Table B.1 Gradation for Georgia PEM Type #7 #789 Granite Granite Screens Lime JMF Control Points % Amount 55 37 7 1 100 Max Min Sive Size Size^0.45 37.5 5.11 100 100 100 100 100 25 4.26 100 100 100 100 100 19 3.76 100 100 100 100 100 100 100 12.5 3.12 82 100 100 100 90 100 80 9.5 2.75 28 99 100 100 60 60 35 4.75 2.02 2 39 99 100 23 25 10 2.36 1.47 2 6 69 100 9 10 1.18 1.08 2 2 46 100 6 0.6 0.79 1 1 30 100 4 0.3 0.58 1 1 17 100 3 0.15 0.43 0 1 7 100 2 0.075 0.31 0 0 1 100 1 4 1 5 102

PAGE 116

103 Table B.2 Bulk Specific Gravity for Georgia PEM AC (%) Number Height (cm) Weight (gms) Bulk Specific Gravity Avg Bulk Specific Gravity 5.5 1 13.668 4659.3 1.930 1.936 2 13.618 4657.1 1.936 3 13.586 4658.3 1.941 6.0 1 13.539 4682.0 1.958 1.961 2 13.557 4683.0 1.956 3 13.468 4681.5 1.968 6.5 1 13.583 4704.9 1.961 1.967 2 13.449 4701.8 1.979 3 13.598 4707.1 1.960 Table B.3 Rice Test for Georgia PEM % A/C 5.5 6 6.5 Wt. Flask+Sample 2876 2867.6 2884.5 2851.7 2892 2892.6 Wt Flask 1872.9 1872.9 1875.7 1844.8 1872.9 1872.9 Wt Sample (A) 1003.1 994.7 1008.8 1006.9 1019.1 1019.7 Wt Flask+Water(D) 6126 6126 6125 6117.6 6126.1 6075.6 Wt Flask+Water+Sample(E) 6719.5 6714.7 6715.8 6707.4 6720.6 6671.4 SSD(B) 1005.4 995.4 1009.2 1007.2 1022.2 1022.2 Multiplier 1.00061 1.00038 1.00095 1.00095 1.00084 1.00084 Gmm 2.437 2.447 2.413 2.415 2.385 2.393 Avg Gmm 2.442 2.414 2.389 % Agg 0.945 0.940 0.935 Gse 2.647 2.660 2.640 2.641 2.625 2.636 Avg Gse 2.641

PAGE 117

104 Table B.4 Drain-down Test for Georgia PEM %AC: 6.0 Mix Type: GPEM Sample: A B M i, (g) Weight of mix before 1-hr aging 1274.2 1275.3 P f, (g) (weight of paper disc + asphalt after draindown) 10.4 10.3 P i (g) (Initial Wt. Of paper Disc) 10.3 10.2 D (%Draindown) 0.01 0.01 D avg (Avg) 0.01 Drain-down test: Passes X Fails

PAGE 118

105 Table B.5 Film Thickness for Georgia PEM Sieve Percent Surface Area Factor SurfaceArea Size Passing ft.2/lb. m2/Kg ft2/lb. m2/Kg 11/2 in.(37.5mm) 100 1 in. (25.0mm) 100 3/4 in. (19.0mm) 100 1/2 in. (12.5mm) 90 3/8 in .( 9.5mm ) 60 2.0 0.41 No. 4 (4.75mm) 23 2 0.41 0.5 0.10 No. 8 (2.36mm) 9 4 0.82 0.4 0.08 No.16 (1.18mm) 6 8 1.64 0.5 0.10 No.30 ( 600um ) 4 14 2.87 0.6 0.12 No.50 ( 300um ) 3 30 6.14 0.9 0.19 No.100 (150um ) 2 60 12.29 1.1 0.23 No.200 ( 75um ) 1 160 32.77 1.7 0.35 hrs Total Surface Area 7.6 1.57 AC % = 6.0 Film Thickness = [ 453.6 g per Pounds divided by % Aggregate ] [ 453.6 g per Pounds ] Surface area in square ft / lb 0. 09290Sq. m per sq. ft. Sp. gr. of AC Or = 453.6 divided by 0.94 Minus 453.6 28.953 divided by 0.731 Or = Film Thickness 39.6 Micron Coating

PAGE 119

106 Effective Sp Gravof Agg% ACGmmGmbVMAVTMVFA2.6415.52.4421.93630.7420.7232.6062.4141.96130.2318.7837.866.52.3891.96730.3817.6841.82Optimum AC6% Mixing Temperature330 F Mineral Fiber0.4% of Total Mix Compaction Temperature325 F VMA30.1030.2030.3030.4030.5030.6030.7030.805.45.65.866.26.46.6% ACVMA VTM17.0018.0019.0020.0021.005.45.65.866.26.46.6% ACVTM VFA30.0032.0034.0036.0038.0040.0042.0044.0055.566.57% ACVMA Fig B.1 Final Mix Design for Georgia PEM

PAGE 120

APPENDIX C IDT TEST RESULTS FOR FC-5, NOVACHIP AND GEOGIA PEM

PAGE 121

108 Table C.1 Stresses Obtained from KENLAYER for ER Calculation Type Mr(Gpa) Mr(Psi) Thickness (in) Stress (psi) Poisson Field Mix-Unaged FC-5 Granite 4.98 722288 0.75 111.9 0.3 FC-5 Limestone 7.35 1066027 0.75 110.7 0.24 NOVACHIP 7.97 1155951 0.75 110.2 0.24 Georgia PEM 4.97 720838 1.25 88.16 0.25 Field Mix-Aged FC-5 Granite 4.81 697051 0.75 112.1 0.3 FC-5 Limestone 7.57 1097936 0.75 110.4 0.25 NOVACHIP 8.71 1263279 0.75 109.7 0.24 Georgia PEM 4.9 710685 1.25 88.23 0.25 Saturated-Unaged FC-5 Granite (25% Red. In Air Voids) 5.31 770150 0.75 111.6 0.3 FC-5 Limestone (25% Red. In Air Voids) 6.17 894883 0.75 111.9 0.23 NOVACHIP (50% Red. In Air Voids) 6.95 1008447 0.75 110.9 0.25 Saturated-Aged FC-5 Granite (25% Red. In Air Voids) 4.35 630914 0.75 112.6 0.3 FC-5 Limestone (25% Red. In Air Voids) 8.18 1186409 0.75 108.3 0.35 NOVACHIP (25% Red. In Air Voids) 6.83 990608 0.75 110.2 0.3 FC-5 Limestone Unaged 7.21 1045722 0.75 110.8 0.24 2 Day Aging 6.89 999310 0.75 110.9 0.24 5 Day Aging 7.41 1074 730 0.75 110.5 0.25 10 Day Aging 8.82 1279 233 0.75 109.5 0.25 Conditioned 8.93 1295 187 0.75 108.7 0.3

PAGE 122

Table C.2 Summary of Fracture Results for Saturated Mix Summary of Mixture Properties OGFC Sample Property 120 ResilientModulus (Gpa) Creep compliance at 1000 seconds (1/Gpa) Tensile Strength (Mpa) Fracture Energy (kJ/m^3) Failure Strain (10-6) m-value D0 D 1 DCSE (kJ/m^3) Energy Ratio Saturated-Unaged Temperature:10 o C FC-5 G (25% Red. In Air Voids) 5.31 9.14 1.41 4.40 3643.1 0.64 1.30E-06 7.08E-07 4.21 2.00 FC-5 L (25% Red. In Air Voids) 6.17 7.14 1.35 1.40 1444.5 0.7115 1.12E-06 3.23E-07 1.25 0.98 NOVACHIP (50% Red. In Air Voids) 6.95 7.23 2.04 9.20 5201.9 0.72 9.92E-07 3.23E-07 8.90 6.23 Saturated-Aged Temperature:10 o C FC-5 G (25% Red. In Air Voids) 4.35 10.63 1.27 2.00 2082.5 0.86 1.59E-06 1.78E-07 1.81 1.46 FC-5 L (25% Red. In Air Voids) 8.18 1.81 1.48 1.00 986.8 0.4522 8.43E-07 5.04E-07 0.87 1.77 NOVACHIP (25% Red. In Air Voids) 6.83 6.35 1.95 5.60 3540.5 0.649 1.01E-06 4.69E-07 5.32 3.55 109

PAGE 123

Table C.3 Summary of Fracture Results for Field Mix and Georgia PEM 110 Sample Property ResilientModulus (Gpa) Creep compliance at 1000 seconds (1/Gpa) Tensile Strength (Mpa) Fracture Energy (kJ/m^3) Failure Strain (10-6) m-value D0 D1 DCSE(kJ/m^3) Energy Ratio FC-5 G 4.98 7.23 1.16 3.00 3247.7 0.5972 1.38E-06 7.84E-07 2.86 1.59 FC-5 L 7.35 1.881 1.11 0.9 982.11 0.4827 9.38E-07 4.29E-07 0.82 1.62 NOVACHIP 7.97 5.916 1.81 4.7 3188.59 0.6214 8.65E-07 5.46E-07 4.49 3.00 GPEM 4.97 19.933 1.24 4.2 4383.23 0.7441 1.39E-06 8.35E-07 4.05 1.95 Field Mix-Aged Temperature: 10 C FC-5 G 4.81 8.744 0.89 1.00 1454.0 0.774 1.43E-06 2.80E-07 0.92 0.68 FC-5 L 7.57 1.808 1.69 2.1 1609.14 0.4301 9.11E-07 5.92E-07 1.91 3.57 NOVACHIP 8.71 4.832 1.89 3.7 2616.75 0.5913 7.92E-07 5.48E-07 3.49 2.69 GPEM 4.9 10.925 0.97 1.1 1552.27 0.6992 1.41E-06 5.86E-07 1.00 0.86

PAGE 124

111 Table C.4 Summary of Fract ure Results for Different Aged Fc-5 Lime Mixes Sample Property Resilient Modulus (Gpa) Creep compliance at 1000 seconds (1/Gpa) Tensile Strength (Mpa) Fracture Energy (kJ/m^3) Failure Strain (10-6) m-value D1 DCSE (kJ/m^3) Energy Ratio FC-5 Lime Temperature: 10 C Unaged 7.21 2.977 1.25 0.8 935.77 0.5892 3.12E-07 0.692 0.85 2 Day Aging 6.89 1.787 1.55 1.6 1473.87 0.4702 4.22E-07 1.426 2.44 5 Day Aging 7.41 1.798 1.54 1.4 1241.35 0.5057 3.34E-07 1.240 2.16 10 Day Aging 8.82 0.683 1.54 0.8 769.51 0.4145 2.29E-07 0.402 1.84 Conditioned 8.93 1.364 1.32 1 1035.58 0.4242 4.51E-07 0.902 2.03

PAGE 125

112 m(Saturated)0.000.100.200.300.400.500.600.700.800.901.00FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)m Unaged Aged Fig C.1 m value for Saturated Field Mix D1(Saturated)0.00E+001.00E-072.00E-073.00E-074.00E-075.00E-076.00E-077.00E-078.00E-079.00E-071.00E-06FC-5 G(25% Red. In AirVoids)FC-5 L(25% Red. In AirVoids)NOVACHIP(50% Red. In AirVoids)D1(1/psi) Unaged Aged Fig C.2 D1 value for Saturated Field Mix

PAGE 126

113 m(Field Mix)00.10.20.30.40.50.60.70.80.91FC-5 GFC-5 LNOVACHIPm Unaged Aged Fig C.3 m value for Field Mix and Georgia PEM D1(Field Mix)0.00E+001.00E-072.00E-073.00E-074.00E-075.00E-076.00E-077.00E-078.00E-079.00E-071.00E-06FC-5 GFC-5 LNOVACHIPD1(1/psi) Unaged Aged Fig C.4 D1 value for Field Mix and Georgia PEM

PAGE 127

114 0.00E+005.00E-081.00E-071.50E-072.00E-072.50E-073.00E-073.50E-074.00E-074.50E-075.00E-07Unaged2 Day Aging5 Day Aging10 Day AgingD1(1/psi) Fig C.5 m value for FC-5 Lime at Different Stages of Aging 00.10.20.30.40.50.60.70.80.91Unaged2 Day Aging5 Day Aging10 Day Agingm Fig C.6 D1 value for FC-5 Lime at Different Stages of Aging

PAGE 128

115 0.390.400.410.420.430.440.450.460.470.480.49UnconditionedConditionedm Fig C.7 m value for conditioned FC-5 Lime 4.15E-074.20E-074.25E-074.30E-074.35E-074.40E-074.45E-074.50E-074.55E-07UnconditionedConditionedD1(1/psi) Fig C.8 D1 value for Aged conditioned FC-5 Lime

PAGE 129

LIST OF REFERENCES 1. Colwill, D.M., G.J Bowskill, J.C. Nichols, and M.E. Daines. Porous asphalt Trials in United Kingdom. In Transportation Research Record 1427,TRB, National research Council, Washington, D.C., 1993, pp. 13 2. Decoene, Y. Contribution of Cellulose Fibers to the Performance of Porous Asphalt. In Transportation Research Record 1265,TRB, National research Council, Washington, D.C., 1990, pp. 82-86. 3. Estakhri,C.K., and J.W.Button. Evaluation of Ultrathin Friction Course. In Transportation Research Record 1454,TRB, National research Council, Washington, D.C., 1994, pp. 9-18. 4. Estakhri,C.K., and J.W.Button. Performance Evaluation of NOVACHIP: Ultrathin Friction Course. Research Report 553-2F, Texas Transportation Institute, College Station, November 1995 5. Heystraeten, G.V., and C. Moraux. Ten Years of Porous Asphalt in Belgium. In Transportation Research Record 1265,TRB, National research Council, Washington, D.C., 1990, pp. 34-40. 6. Huber, G. Performance of Open Graded Friction Course Mixes. In NCHRP Synthesis of Highway Practice 284, TRB, National Research Council, Washington, D.C., 2000 7. Huddlestone, I.J., H. Zhou, and R.G Hicks. Evaluation of Open-Graded Asphalt Concrete Mixtures Used in Oregon. In Transportation Research Record 1427,TRB, National Research Council, Washington, D.C., 1993, pp. 5-12. 8. Huet, M., A. Boissoudy, J. Gramsammer, A. Bauduin, and J. Samanos. Experiments with Porous Asphalt on the Nantes Fatigue Test track. In Transportation Research Record 1265, TRB, National Research Council, Washington, D.C., 1990, pp. 54-58. 9. Jimenez, F.E., and J. Gordillo. Optimization of Porous Mixes through the Use of Special Binders. In Transportation Research Record 1265,TRB, National Research Council, Washington, D.C., 1990, pp. 59-68. 10. Kandhal, P.S., and L. Locket. Construction and Performance of Ultrathin Friction Course. NCAT Report 97-5, Auburn University, September 1997 116

PAGE 130

117 11. Knoll, E., and E.J. Buczeskie. Ultrathin Hot Mix Asphalt Resurfacing (Hot Laid Microsurfacing). Research Report 93-067, Pennsylvania Department of Transportation, February 1999. 12. Rajib, B. M., P. S. kandhal, L. A. Cooley, and D. E. Watson. Design, Construction and Performance of New Generation Open-Graded Friction Courses. NCAT Report 2001-01, Association of Asphalt Paving technologists, April 2000 13. Roque,R., & W.G. Butlar. The Development of a Measurement and Analysis System to Accurately Determine Asphalt Concrete Properties Using the Indirect Tensile Mode. In Journal of the Association of Asphalt Paving technologists, Vol. 61,1992, pp. 304-332 14. Roque,R., & W.G. Butlar. Development and Evaluation of the Strategic Highway Research Program Measurement and Analysis System for Indirect Tensile Testing at Low temperatures. In Transportation Research Record 1454, TRB, National Research Council, Washington, D.C., 1994, pp. 163-171. 15. Roque,R., Z. Zhang & B. Shankar. Determination of Crack Growth Rate Parameters using the Superpave IDT. In Journal of the Association of Asphalt Paving Technologists, Vol.68, 1999, pp. 404-433. 16. Ruiz, A., R. Alberola, F. Perez, and B. Sanchez. Porous Asphalt Mixtures in Spain. In Transportation Research Record 1265, TRB, National Research Council, Washington, D.C., 1990, pp. 87-94. 17. Serfass, J.P., P. Bense, J. Bonnot, and J. Samanos. New Type of Ultrathin Friction Course. In Transportation Research Record 1304, TRB, National Research Council, Washington, D.C., 1991, pp. 66-72. 18. Seshadri, M. Novachip Paving in Mississippi. Final Report, Mississippi Department of Transportation, August 1993. 19. Smith, R.W., J.M. Rice and S.R. Spelman. Design of Open-Graded Asphalt Friction Courses. Report FHWA-RD-74-2, Federal Highway Administration, January 1974. 20. Vavrik, W.R, S.H. Carpenter. Calculating Air Voids at Specified Number of Gyrations in Superpave Gyratory Compactor. In Transportation Research Record 1630, TRB, National Research Council, Washington, D.C., 1998, pp. 117-125. 21. Watson, D., A. Johnson, and D. Jared. Georgia Department of Transportations Progress in Open Graded Friction Course Development. In Transportation Research Record 1616,TRB, National research Council, Washington, D.C., 1998 pp. 30-33.

PAGE 131

BIOGRAPHICAL SKETCH Arvind Varadhan was born on September 3, 1980, in the city of Bombay, India. He received his bachelors degree in civil engineering from Indian Institute of Technology, Delhi, in August 2002. After his undergraduate studies, He came to the University of Florida to pursue a Master of Engineering degree. He plans to work in a geotechnical engineering consultancy firm in Florida after he graduates with his M.E. degree. 118


Permanent Link: http://ufdc.ufl.edu/UFE0006043/00001

Material Information

Title: Evaluation of Open-Graded and Bonded Friction Course for Florida
Physical Description: Mixed Material
Copyright Date: 2008

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: UFE0006043:00001

Permanent Link: http://ufdc.ufl.edu/UFE0006043/00001

Material Information

Title: Evaluation of Open-Graded and Bonded Friction Course for Florida
Physical Description: Mixed Material
Copyright Date: 2008

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: UFE0006043:00001


This item has the following downloads:


Full Text












EVALUATION OF OPEN-GRADED AND BONDED FRICTION COURSE FOR
FLORIDA

















By

ARVIND VARADHAN


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


2004















ACKNOWLEDGMENTS

I would like to thank Dr. Bjorn Birgisson and Dr. Reynaldo Roque for their

guidance throughout the project. I believe that their knowledge and expertise helped me

understand and solve crucial problems during the course of the project.

I really appreciate the technical support and advice I received from Georg Lopp

throughout my research work. I would like to thank Lokendra, Claude, Tanya and Eddy

for their assistance in performing various laboratory tests. I would also like to thank

Christos Drakos for helping me with KENLAYER in performing stress analysis.

I would like to thank Greg Sholar, Shanna Jhonson and Ricky Lloyd from the

research wing of DOT for their help during the course of the project.

I would like to thank all my friends for providing an unforgettable and enjoyable

time during my two years of study in Gainesville. Finally, I would like to thank my

parents and my brother for all the love and support they given me throughout my

academic years.
















TABLE OF CONTENTS
Page

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

LIST OF TABLES ............... ...................... ...................... ...... .. ............. vi

LIST OF FIGURES .................. .............................................. ..............viii

ABSTRACT................................. .............. xii

CHAPTER

1 INTRODU CTION .................. ...........................................................1

1.1 Thick Open Graded Friction Courses....................... .......................................1
1.2 Bonded Friction Course Evaluation.............................. ...............2
1.3 Objective..................................... .................. ............. .........3
1.3.1 FC-5 and Novachip............... ........ ...................3
1.3.2 Georgia PEM ................... .................................... ... .... ............ .4

2 LITERATURE REVIEW .................................................. ...............5

2. 1 NovachipNovachip D..........escription ..........................................................................5
2.1.1 Novachip Description.........................5
2.1.2 P having E quipm ent ........................................ .................6
2.1.3 Post Construction Testing..................... ... ..................
2.1.3.1 Surface roughness International roughness index...........................7
2.1.3.2 Skid friction ...... ......... ......... .... .. ....... .7
2.1.3.3 Surface m acrostructure............................ ...............8
2.1.3.4 Ride quality data......................... ......... .... ..... .. 8
2.1.3.5 Rolling Noise................................... ........ .... ....... .9
2.1.4 NOVACHIP M ix Design ................................ ...............9
2.1.4.1 Gradation ............... .. ....... ..... .. .........9
2.1.4.2 Asphalt content determination ................. ................. ............9
2.1.5.1 Surface treatm ent paving m machine ................................................. 11
2.1.5.2 Rollers .............. .......... .............. .... .......... 13
2.1.5.3 Straightedges and tem plates ........................................ .................13
2.1.5.4 W weather lim stations ........................................ ............ ... ....13
2.1.5.5 Tack coat ............. ........... ................ ....... ........13
2.1.5.6 Hauling equipment ...................... .................. 14
2.1.5.7 Spreading and finishing............................... .... ........ 14










2.1.5.7 C om action ............................................... .............. 14
2.1.5.8 Method of measurement............. ...........................15
2.2 Porous European M ix (PEM ) ........................................................ 15
2.2.1 A advantages of Porous A sphalt .............................................. ......15
2.2.1.1 Hydroplaning and glare reduction.........................15
2.2.1.2 Noise reduction .......... .......................... 16
2.2.1.3 Skid friction ............... .... ............ .... ..............17
2.2.2 Disadvantages of Porous Asphalt.................................17
2.2.2.1 Strength .................................... ..................... 17
2.2.2.2 Initial stiffness m odulus .............................................. ......18
2.2.2.3 Aging and stripping............................ .......... 18
2.2.2.4 T em perature ............................................................. 18
2.2.2.5 Clogging ........... .......... ................. ........19
2.2.3 Performance Related Laboratory Testing........... ........ ............19
2.2.3.1 Resistance to plastic deformation................................. 19
2.2.3.2 Resistance to indirect tension.................................20
2.2.3.3 Resistance to disintegration........................ ............. 20
2.2.3.4 Adhesiveness ............... ............. ........20
2.2.3.5 D rainage test.................................... ...................21
2.2.4 M ix D esign A approach ........................................ ................. 21
2.2.4.1 British design........... .... ....................................22
2.2.4.2 Spanish design ....... .... ........ ....................... 22
2.2.4.3 Italian design .............................................. .............. 23
2.2.4.4 B elgium design.................................. .................... 23
2.3 Georgia PEM ............................. .......................... ........ 24
2.3.1 M material Selection................ .... .... ............... 23
2.3.1.1 C om position ............................... ........... .............. 23
2.3.1.2 Gradation.................... .. ......... .............26
2.3.1.3 Aggregate specification......................... .............. 27
2.3.1.4 Polym er m odified asphalt cem ent................................................ 27
2 .3.1.5 M ineral fibers ............ .............. .............................. ............... 28
2.3.2 Georgia PEM mix design procedure ...................................... 28

3 MATERIAL S ............................. ..................31

3.1 FC-5 and Novachip ............... .................. ......31
3.1.1 Aggregates................ .. ......... ........ ........31
3.1.2 Binder ............. ...................... .........31
3.2 Georgia PEM ................. .. ............................ 32
3.2.1 Aggregate and Binder ................................. ............... 32

4 DETERMINATION OF COMPACTION LEVEL AND AGING PROCEDURE FOR
FRICTION COURSES ............................................................................. 34

4.1 Compaction of the Friction Course Mix...................................................34
4.1.1 Compaction Data ................ ......................................... .. ...... 34
4.1.2 Initial Study ...................................... ............ ........ 38










4.2 A sphalt Content D eterm nation .................................................................... 44
4.3 M ixing and Com action ............................................. ............... 45
4.4 Long-Term Oven Aging of Friction Course ..................................................47

5 INDIRECT TENSILE TESTING.......................................................... ... ...... 49

5.1 Sam ple Preparation ............................ .... ........................... 50
5.2 Test Procedures......................... ....... .. ..........53
5.2.1 R esilient M odulus Test....................................... ............... 53
5.2.2 Creep Test.................................... .........55
5.2.3 Strength Test ................ .............. .... .. ......... .. .. ............... 56
5.3 Issues Related with IDT Testing of Friction Course ............... ...............59

6 MIX DESIGN APPROACH FOR GEORGIA PEM.................................................61

6.1 Evaluation of Compaction Level for Georgia PEM ........................................ 61
6.2 M ix D esign A pproach................................................... 61

7 FINDINGS AND ANALYSIS ................................................65

7.1 Evaluation of FC-5 and Novachip Field Mix .......................................... 66
7.1.1 Unaged Field Mix.................................... ........67
7.1.2 Aged Field M ix.......................................... ..... ...67
7.1.3 Saturated U naged M ix ................................................................... 74
7.1.4 Saturated A ged M ix.................................... ................... 75
7.1.5 Moisture Conditioning.................................... 83
7.2 Evaluation of G eorgia PEM .................................... .................. 88

8 SUMMARY, FINDINGS AND ANALYSIS....... ..........................................94

8.1 Sum m ary of F finding s .................................................................................... 94
8.2 Conclusions............................. .... ........ .. .95
8.3 Recommendations....................... .. .............96

APPENDIX

A VOLUMETRICS FOR FC-5 AND NOVACHIP............................97

B VOLUMETRICS FOR GEORGIA PEM .............. .............. ...............101

C IDT TEST RESULTS FOR FC-5, NOVACHIP AND GEOGIA PEM .................107

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

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
















LIST OF TABLES

Table Page

2.1 International Roughness Index ................. .................................7

2.2 Skid Number................................... .......... ...............8

2.3 Ride Quality .......................... ............ .............. .........9

2.4 JMF Gradation Range .................. .............................10

2.5 G eorgia PEM G radiation .................................................. ............... 26

2.6 Aggregate Specifications for Georgia PEM .........................................27

2.7 Design Requirements for Georgia PEM ..........................................28

3.2 Mixes from the Field..................... .................32

4.1 A ir V oids ..................................................35

4.2 Locking point For the Mixtures ................................ ....................... ......38

4.3 Locking Point Based on Gradient of Slope...............................................................41

4.4 Locking Points of all Mixtures based on Gradient of Slope......................................41

4.5 A ir V oids for 50 G yrations ............................................... ............... 42

4.7 W eight of PG 76-22 to be added.................. .........................................................47

6.1 Locking Point for Georgia PEM ........................................................... .. ...... 62

A. 1 Gradation for FC-5 Granite, Limestone and Novachip............... ............ 98

A.2 Rice Test for the Friction Courses ................................. ............... 99

A.3 Air Voids and Asphalt Content for Friction course..................................................100

B. 1 Gradation for Georgia PEM ....................................................................102

B.2 Bulk Specific Gravity for Georgia PEM ............................................................... 103










B .3 R ice Test for G eorgia PEM .......................................................................... ....... ..... 103

B.4 Drain-down Test for Georgia PEM........................................................... 104

B.5 Film Thickness for Georgia PEM ................................ ............... 105

C.2 Summary of Fracture Results for Saturated M ix ......................................................109

C.3 Summary of Fracture Results for Field Mix and Georgia PEM ..............................110

C.4 Summary of Fracture Results for Different Aged Fc-5 Lime Mixes ...............1.........11
















LIST OF FIGURES

Figure page

2.1 Surface Constant Kc vs. Percentage Oil Retained.....................................................12

2.2 British Gradations ....................................... ........ .. .... ..............24

2.3 Spanish Gradation........................ .... ....... .........25

2.4 Italian Gradation Band.............................................................................26

2.5 Georgia PEM Gradation Band.............................................................. ............ 27

3.1 Gradation of FC-5 Granite, FC-5 Limestone, Novachip and Georgia PEM ............33

4.1 Compaction Curve for FC-5 Limestone ..........................................36

4.2 Compaction Curve for NOVACHIP................................ .................... ......36

4.3 Compaction for FC-5 Granite ............. .................................. .......37

4.4 Locking Point for FC-5 with Limestone by Visual Observation............... ..............39

4.5 Locking Point for Novachip by Visual Observation..................................................39

4.6 Locking Point for FC-5 with Granite by Visual Observation................ ...............40

4.7 G radations after Extraction .............................................. ............... 43

4.8 1/8 M esh used for Containing the Pill................................................... 48

4.9 Pill Contained w ith M esh.......................................................... 48

5.1 Effect of Loading Condition .............................................................. .............49

5.2 Cutting D vice ...................................... .............................. ........ 51

5.3 G auge P point A ttachm ent .............................................................................................52

5.4 M parking L loading A xes ............................................................53

5.5 Power M odel for Creep Compliance ............................... ...............57



viii









5.6 FE and DCSE from Strength Test.................................... .................. 59

6.1 Final M ix Design for Georgia PEM .................................. ...... ............... 64

7.1 Typical Pavement Structure for Stress Calculation ......................................... 66

7.2 Fracture Energy for Field M ix .............................................................. ............ 68

7.3 Dissipated Creep Strain Energy for Field Mix ................. ................. ............68

7.4 E energy R atios for F field M ix .................................................................................. 69

7.5 Failure Strain for Field M ix ...........................................................................69

7.6 Tensile Strength for Field Mix.............................................70

7.7 Resilient M odulus for Field M ix.................................... ................... 70

7.8 Creep R ate for Field M ix ....................................................................................... ....71

7.9 Aging Effect in Fc-5 Granite and FC-5 Limestone ......................................73

7.10 FE for FC-5 Lime at Various Stages of Aging ..................................75

7.11 DCSE for FC-5 Lime at Various Stages of Aging........................76

7.12 Failure Strain for FC-5 Lime at Various Stages of Aging......................76

7.13 Energy Ratio for FC-5 Lime at Various Stages of Aging ............... ...................77

7.14 Tensile Strength for FC-5 Lime at Various Stages of Aging..............................77

7.15 Resilient Modulus for FC-5 Lime at Various Stages of Aging .............................78

7.16 Creep Rate for FC-5 Lime at Various Stages of Aging .................. .... ...........78

7.17 Comparison of ER between unaged Saturated and Field Mix...............................79

7.18 FE for Saturated M ix .............................................................79

7.19 DCSE for Saturated Mix.................... ..... ....... ........80

7.20 Failure Strain for Saturated M ix ........................................................ 80

7.21 ER for Saturated Mix .............. .................. ..........81

7.22 Tensile Strength for Saturated Mix.................. .............. .......... 81

7.23 Resilient M odulus for Saturated M ix................................ ............... 82









7.24 Creep Strain Rate for Saturated M ix...........................................82

7.25 Creep Deformation Curve for Conditioned samples of FC-5 with Granite............84

7.26 Creep Rate for Unconditioned and Conditioned samples FC-5 Granite.................85

7.27 FE after Conditioning for FC-5 with Limestone...................................................85

7.28 DCSE after Conditioning for FC-5 with Limestone..........................................86

7.29 ER after Conditioning for FC-5 with Limestone........... .......... ..............86

7.30 Failure Strain after Conditioning for FC-5 with Limestone .....................................87

7.31 Mr after Conditioning for FC-5 with Limestone ................................... ........87

7.32 Creep Rate after Conditioning for FC-5 with Limestone .....................................88

7.32 Loosely and Well Reinforced Matrix of Aggregate and Asphalt .............................89

7.33 FE for Georgia PEM ........................................90

7.34 D CSE for Georgia PEM ............................................. ............... 90

7.35 Failure Strain for Georgia PEM ......................................................... .. ......91

7.36 ER for G eorgia PEM .........................................................91

7.37 Tensile Strength for Georgia PEM ................................ ............... 92

7.38 Resilient Modulus for Georgia PEM ............................. ...............92

7.39 Creep R ate for G eorgia PEM ........................................................................................93

B. 1 Final M ix Design for Georgia PEM ................................................. ........... 106

C. 1 m value for Saturated Field M ix ...................................... ................ ..... 112

C .2 D l value for Saturated Field M ix......................................................... .... ... 112

C.3 m value for Field Mix and Georgia PEM........................................................ 113

C.4 Dl value for Field Mix and Georgia PEM.............................................................113

C.5 m value for FC-5 Lime at Different Stages of Aging .... ........... .................114

C.6 Dl value for FC-5 Lime at Different Stages of Aging................ ......... ..........114

C.7 m value for conditioned FC-5 Lime................................... ...................115










C.8 Dl Value for Aged conditioned FC-5 Lime.......................................................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

EVALUATION OF OPEN-GRADED AND BONDED FRICTION COURSE FOR
FLORIDA

By

Arvind Varadhan

August 2004

Chair: Bjorn Birgisson
Cochair: Reynaldo Roque
Major Department: Civil and Coastal Engineering

The project involves the fracture evaluation of 1) FC-5 with Granite and limestone

(Friction course used in Florida) and NOVACHIP (bonded friction course) obtained from

the field, and 2) Georgia PEM prepared in the lab. As a part of the study, an extensive

literature review was done covering the various kinds of friction courses used around the

world. The focus of literature review was on various mix design approaches for thick

open graded and bonded friction courses and materials, equipment and construction

guidelines as specified in different states in US.

The field mixes obtained from the US 27 Highway Project were tested in the

laboratory to evaluate their fracture properties. In addition, the effect of aging and

moisture damage on their fracture resistance was also evaluated.

The first step in the process was the identification of the compaction level for

these friction courses. Thus, the locking point for the friction course was identified based

on our study of the rate of change of compaction.









To replicate the interface, consisting of Novabond, between the underlying HMA

and the friction course, it was decided that PG 76-22 should be added to the field mix

since the tack coat essentially contained PG 76-22. However, there were problems related

to the workability of the mixes at such high asphalt content, which were solved by

determining the compaction and mixing temperature for these mixes.

Finally, the fracture testing was done using Superpave IDT test and the fracture

properties were evaluated using the framework of HMA fracture mechanics. The issues

concerning the IDT testing on the friction course were also identified and dealt with.

In the case of Georgia PEM, the primary objective was to compare its fracture

performance with FC-5 limestone and granite. The Georgia PEM mixes were prepared

using the same mix design as used by Georgia DOT. Both unaged and aged mixes were

evaluated for their fracture performance using IDT testing and the results were compared

with FC-5 and Novachip.














CHAPTER 1
INTRODUCTION

This project focuses on two different open graded friction course approaches:

The first one is the evaluation of thick open graded friction courses for Florida

conditions, and the second one is the evaluation of bonded friction courses in Florida. In

the following the motivation and background for studying each of these materials has

been provided and the objectives of the research project have been outlined.

1.1 Thick Open Graded Friction Courses

Because of the frequency of high intensity short-duration rainfall events in Florida,

vehicular hydroplaning is a serious concern, especially at the Interstate and other limited

access facilities. Along with pavement cross-slope and rutting, the surface texture of an

asphalt pavement plays a critical role in the prevention of hydroplaning on high-speed,

multi-lane facilities. In order to minimize problems of this nature, a number of states

(including Florida) have utilized traditional open-graded friction courses in order to

maximize the pavement's macro texture, which in turn reduces splash and spray, and

minimizes hydroplaning potential.

In the mid-1970, Florida developed a friction course (FC-2) which was a 3/8-inch

Nominal Maximum Aggregate Size (NMAS) open-graded mixture (with polish resistant

aggregate) placed approximately 1/2-inch thick. In the 990's, the FC-2 was eventually

replaced by a slightly coarser open-graded friction course (FC-5), which is a 1/2 -inch

NMAS mixture, placed approximately % inch thick.









Since the FC-5 mixture is coarser and is placed slightly thicker than the FC-2, it has

a greater capacity to store/drain water from the pavement surface during a severe

rainstorm. However, this additional storage capacity has never been quantified. Visual

observations of this surface type indicate that the while it is better than the old FC-2, it

will still "fill-up" with water resulting in water ponding on the pavement surface.

A number of European countries, as well as several states in the US (Georgia,

Oregon) have developed a "porous friction course" which is an open-graded friction

course placed in thicknesses ranging from 11/4 to 2 inches thick. The combination of the

high in-place air void content of this mixture, coupled with the thickness at which it is

placed, gives this type of pavement surface a great deal of potential with regard to storage

and drainage of water during a rainstorm. However, there are also a number of questions

associated with these types of pavements, such as fracture resistance, rutting resistance,

and long-term porosity.

1.2 Bonded Friction Course Evaluation

There is a high priority need to evaluate the estimated performance life and cost

effectiveness of a bonded friction course that uses a special paver to lay a heavy polymer-

modified tack coat just in front of the hot mix mat. This technology is available in

Florida and has been demonstrated on several small projects. Other states, including

Texas, Pennsylvania, and Alabama have several years experience with this process and

indications are that it has great potential.

The ongoing Longitudinal Wheel path Cracking study at UF has confirmed that a

significant amount of the distress on high-volume Florida pavements is caused by surface

initiated wheel path cracking from lateral stresses generated by radial truck tires. A

recent evaluation of the ten year performance of ground tire rubber in an open graded









friction course on SR 16 in Bradford County indicates that increased asphalt content from

ground tire rubber modification of the friction course can reduce longitudinal wheel path

cracking. Based on these studies, there is a high likelihood that the bonded friction

course process could significantly extend the crack resistance life of open graded mixes

in Florida by providing more polymer modified asphalt to the friction course.

1.3 Objective

This study involves the fracture evaluation of 1) FC-5 (Friction course used in

Florida) and Novachip (bonded friction course) obtained from field, and 2) Georgia PEM

prepared in lab. As a part of the study, an extensive literature review was done covering

the various kinds of friction courses used around the world. The focus of literature review

was on various mix design approaches for thick open graded and bonded friction courses

and materials, equipment and construction guidelines used by the different states.

1.3.1 FC-5 and Novachip

In July 2003, five test sections were laid on US 27 highway in Highlands County of

South Florida. These test sections were: 1) FC-5 Limestone, 2) FC-5 Limestone with

Novabond, 3) FC-5 Granite, 4) FC-5 Granite with Novabond, and 5) NOVACHIP. The

field mixes of the above test sections were tested in the laboratory to evaluate their

fracture properties. In addition, the effect of moisture conditioning on the fracture

resistance was also evaluated for FC-5.

The first step was to identify the compaction level for these friction courses.

Hence, it was important to identify the locking point for these mixes, which is defined as

the point beyond which the resistance to compaction increases significantly. The locking

point for the friction course was identified based on our study of the rate of change of

compaction.









To replicate the interface, consisting of Novabond, between the underlying HMA

and the friction course, it was decided that PG 76-22 be added to the field mix since the

tack coat essentially contained PG 76-22. However, there were problems related with the

workability of the mixes at such high asphalt content, which were solved by determining

the compaction and mixing temperature for the mixes.

Finally, the fracture testing was done using Superpave IDT testing and the

fracture properties were evaluated using the framework of HMA Fracture Mechanics.

The issues concerning the IDT testing on the friction course were also identified and dealt

with.

1.3.2 Georgia PEM

Georgia PEM is a kind of Porous European Mix adopted by Georgia DOT. The

primary difference between normal FC-5 and Georgia PEM is the use of polymer

modified asphalt and addition of mineral fiber to the mix. .

Georgia PEM mixes were prepared using the same mix design as used by Georgia

DOT. However, Superpave Gyratory Compactor was used instead of Marshall blows for

compaction. Both unaged and aged mixes were evaluated for their fracture performance

using IDT testing and the results were compared with FC-5 and Novachip.














CHAPTER 2
LITERATURE REVIEW

An extensive literature review was done covering the mix design, construction and

performance of various kinds of friction courses. The main focus of literature review was

on 1) various mix design approaches for thick open graded and bonded friction courses,

2) materials, equipment and construction guidelines used by the different states in US.

The literature review also covered set of guidelines and protocols for producing

representative laboratory mixtures for evaluation and testing, as well as guidelines for the

construction of bonded friction courses (Novachip).

2.1 Novachip

Novachip is an ultra thin friction course whose primary objective is to restore the

skid resistance and surface impermeability. The few other advantages are excellent

adhesion, reduced rolling noise, reshaping of existing pavements.

2.1.1 Novachip Description

Novachip consists of a layer of hot precoated aggregate over a binder spray

application. The tack coat is generally a polymer modified, emulsified asphalt (usually a

latex or elastomer modified emulsion. Such a coating offers strong bonding between the

chippings. Thus due to the immediate application of the binder, chippings are perfectly

held in position and whip-off is totally eliminated.

The hot mix material is a gap graded mixture that includes large proportion (70-

80%) of single sized crushed aggregate, bound with mastic composed of sand, filler and

binder. The binder content varies usually from 5-6 percent. The course thickness varies









from 10-20 mm depending on maximum size of stone. Layer thickness is usually 1.5

times the diameter of the largest stone.

Novachip is placed with a specially designed paving machine that combines the

function of asphalt distributor and a lay down machine. The paver applies the tack coat

and the hot mixture in a single pass. This heavy application of tack helps to ensure

adhesion of the friction course to the underlying pavement and to prevent the possibility

surface water from permeating into the pavement. The operation of the paving equipment

has been described in detail by Colwill et al. (1) and Serfass et al. (18).

2.1.2 Paving Equipment

The operation involved in the Novachip process is as follows

* Collection of mixture from the transport truck
* Storage of mixture
* Storage of sufficient tack emulsion for at least 3 hrs of operation
* Distribution of tack coat with servo controlled application rate
* Immediate covering of tack with the mixture
* Smoothing of applied mixture
The Novachip machine includes the following components:
* A receiving hopper for precoated chippings with self locking hook for the
supplying truck
* A scraper-type conveyer
* A chipping storage chamber with appropriate thermal insulation and a total
capacity of 5 m3
* Several binder tanks, thermally insulated, with a capacity of 12 m3 (more than half
day work)
* A conveyer transferring chippings to the screed unit
* A variable-width spray bar
* A variable width heating screed unit

The design of spray bar includes wide-angle nozzles whose delivery depends on the

road speed of the machine. The friction coarse is applied at high speed, > lOm/min and

reaches 20-25 m/min.









2.1.3 Post Construction Testing

Performance evaluation of NOVACHIP has been typically in terms of

measurement of skid resistance, surface macrostructure, surface roughness, and ride

quality data and noise level. The procedure for above tests are described as follows:

2.1.3.1 Surface roughness International roughness index

The international Roughness Index (IRI) measures pavement roughness in terms of

the number of inches per mile that a laser, mounted in a specialized van, jumps as it is

driven across the interstate and expressway system (10). The lower the IRI number, the

smoother the ride. The rating system scores a roadway from the following criteria:

Table 2.1 International Roughness Index

FHWA Interstate National Non-NHS All Other
Highway Routes Highway Traffic Routes & Routes
Statistics System (NHS) Other Routes
Categories Non with
(Inches per mile) Interstate ADT>= 2000
Routes
<60 Excellent Excellent Excellent Excellent
60-94 Good Good Good
95-119 Fair Fair Good
120-144 Fair
145-170 Fair
171-194 Poor Poor
195-220 Poor
>220 Poor


The test is conducted in accordance with ASTM E-950 test method.

2.1.3.2 Skid friction

Pavement skid friction is tested in accordance with ASTM E-274 test method

(locked wheel skid trailer) (10). The classes of skid numbers are as follows









Table 2.2 Skid Number

Skid Number Description
50-100 -Very Good
40-49 Good
30-39 Fair
20-29 Poor
1--19 -Very Poor


2.1.3.3 Surface macrostructure

Pavement friction depends on both microstructure and macrostructure.

Microstructure refers to detailed surface characteristics of the material. Good

microstructure will provide effective contact area between the tire and the aggregate on

the road surface. Macrostructure refers to the general coarseness of the surface, which

facilitates the drainage of water from the surface. Pavement macrostructure is defined as

the deviation of the pavement from a true planar surface and the average texture depth

between the bottom of the pavement surface and the top of surface aggregates.

Surface macrostructure depth is measured by Sand Patch Depth (SPD) volumetric

technique method in accordance with ASTM E 965-87 (10).

2.1.3.4 Ride quality data

Texas DOT uses an instrument called Slometer to measure the ride quality of the

pavement surface (3,4). The ride quality essentially indicated the smoothness or

irregularities and the ruts in the pavements. A Slometer has an accelerometer processing

computer and a data storage computer mounted in one vehicle. The data is converted into

a ride score based on a user panel rating than ranges from 0.1 (very rough) to 5 (very

smooth). The ride score classifications are shown in the table as follows:









Table 2.3 Ride Quality

Ride Score Description
4.0-5.0 Very Smooth
3.0-3.9 Smooth
2.0-2.9 Medium Rough
1.0-1.9 Rough
0.1-0.9 Very rough


2.1.3.5 Rolling noise

Rolling noise measurement can be made using a French-German method i.e using

three different vehicles with different types of tires at various speeds (18). This method

establishes a correlation between acoustical pressure (in dB) and vehicle speed.

2.1.4 NOVACHIP Mix Design

The NOVACHIP mix design is based on the design developed by the FHWA office

of research (19). The material requirement, aggregate gradations are established and the

bulk and apparent specific gravity of coarse and fine aggregate are determined. The

specific gravity of asphalt cement is also obtained. After these preliminary steps, surface

capacity (K,) of the coarse aggregate is calculated and it is used to then calculate the

design asphalt content.

2.1.4.1 Gradation

There are three kinds of gradation band based on the nominal maximum size (9,10).

These design gradation bands for the various nominal maximum sizes are as follows

2.1.4.2 Asphalt content determination

The surface capacity K, is determined by the Centrifuge Kerosene Equivalent

(C.K.E) test based on FHWA design procedure (20). The Kc value is a measure of

relative roughness and degree of porosity of the aggregate and is used in an experienced

based formula to calculate the design asphalt content.









Table 2.4 JMF Gradation Range


Sive Sive Percentage Passing
Size (mm) Size^0.45
A B C
(6.3 mm) (9.5 mm) (12.5 mm
Max Min Max Min Max Min
19 3.76 100 100
12.5 3.12 100 100 100 85
9.5 2.75 100 100 100 75 90 70
6.3 2.29 100 75 45 30 45 30
4.75 2.02 60 40 37 24 40 24
2.36 1.47 24 20 26 21 25 21
1.18 1.08 20 15 23 15 25 18
0.6 0.79 15 10 15 12 20 12
0.3 0.58 12 8 14 8 16 8
0.15 0.43 10 7 10 5 10 5
0.075 0.31 7 5 7 5 7 4
Asphalt Content % 5.3 % Min 6 % Max


Equipment.

* Pans, 115 mm x 25 mm deep.
* Hot plate or oven capable of 1100 C 5' C (230 F 9 F)
* Beaker, glass 1500 ml
* Balance, 500 g capacity,
* Metal funnels, top dia 89 mm, height 114 mm, Orifice 13 mm with piece of 2mm
sieve soldered to the bottom of the opening
* Oil, S.A.E No. 10 lubricating
Surface Capacity (KJ) Test for Coarse Aggregate.

1. Quarter out 105 g of aggregate representative of material passing through 9.5 mm
sieve and retained on 4.75 mm (No. 4) sieve.
2. Dry sample on hotplate or inl 10 C 50 C (230 0 F 9 0 F) oven to a constant mass
and allow to cool
3. Weigh out approximately 100 g sample to nearest 0.1 g and place it in a funnel
4. Completely immerse sample in S.A.E No. 10 lubricating oil for 5 minutes.
5. Drain sample for 2 minutes
6. Place funnel containing sample in 600 C (1400 F) oven for 15 minutes of additional
draining
7. Pour sample from funnel into a tarred pan, cool, reweigh sample to nearest 0.1 g.
Determine the amount of oil retained as percent of dry aggregate mass









8. Use the chart as shown in Fig. 2.1. If the apparent specific gravity for the fraction is
greater than 2.70or less than 2.60 apply correction to percentage oil retained using
the formula as shown in Fig. 2.1.

Design Asphalt Content.

1. If the apparent specific gravity of the coarse aggregate fraction is 2.6 or more but
not greater than 2.7 then
2. AC = 2 Kc + 4.0
3. If the apparent specific gravity of the coarse aggregate fraction is greater than 2.70
or less than 2.60 the
4. AC = (2 Kc + 4.0) 2.65/ Sf
5. AC = Design asphalt content, present by mass of aggregate
6. Sf = Apparent specific gravity of coarse aggregate

Thus this the optimum asphalt content used for final design.

2.1.5 Construction Requirements

The construction details of NOVACHIP as described in Mississippi DOT

specification (19) is as follows:

2.1.5.1 Surface Treatment Paving Machine

Screed unit. The machine shall be equipped with a heated screed. It shall produce

finished surface meeting the requirements of the typical cross section.

Extensions added to the screed shall be provided with the same heating capability

as the main screed unit, except for use on variable width tapered and/ or as approved by

the Engineer. The screed with extensions if necessary shall be of such width as to pave an

entire lane in a single pass.

Asphalt distribution system. A metered mechanical pressure sprayer shall be

provided on the machine to accurately apply and monitor the rate of application of the

tack coat. The rate shall be uniform across the entire paving width. It shall be applied at a

temperature of 1600 F 20 0 F.











% oil ret. corrected = % oil rel. x sp. gr. of aggregate
2.65

3.0
2.8 ----_
2.4 ------____
2.2
o 2.0


Cd
1,4_ __ -------------_ _- --- -- __ __ _
o 1,4



1. 0





1.5 2 3 4 5 6 7'8 9

Per cent oil retained (Corrected for sp. gr. of aggregate)

Fig.2.1 Surface Constant Kc vs. Percentage Oil Retained

Application can be immediately in front of the screed unit. The application rate

shall be 0.22 0.05-gallons/sq. yards, unless otherwise directed by the Engineer.The

application rate shall be verified by the carpet tile test before the work commences. At the

end of each workday, a check shall be made to determine the quantities of tack coat used.

The check shall be made by means of calibrated load cells on the machine

Tractor unit. The tractor unit shall be equipped with a hydraulic hitch sufficient in

design and capacity to maintain contact between the rear wheels of the hauling equipment

and the pusher rollers of the finishing machine while the paving is being unloaded.

The machine shall support no portion of the weight of the hauling equipment, other

than the connection. No vibrations or other options of the hauling equipment, which









could have a detrimental effect on the riding quality of the completed pavement, shall be

transmitted to the machine.

The use of any vehicle which requires dumping directly into the finishing machine

and which the finishing machine cannot push or propel to obtain the desired lines and

grades without resorting to hand finishing will not be allowed.

2.1.5.2 Rollers

Steel wheel rollers shall meet be rated at 10 tons and may be three wheel type but

the tandem type is preferred.

2.1.5.3 Straightedges and templates

When directed by the engineer, the contractor shall provide acceptable 10-foot

straightedges for surface testing.

2.1.5.4 Weather limitations

The tack coat and the paving mixture shall be placed only when the temperature of

the surface to be overlaid is no less than 500 F and rising, but shall not be placed when

the air temperature is below 60 O F and falling. It is further understood that the tack coat

or paving mixture shall be placed only when the humidity, general weather conditions

and moisture conditions of the pavement surface are suitable in the opinion of the

engineer.

2.1.5.5 Tack coat

Before the tack coat and the paving mixture are applied, the surface upon which

the tack coat is to be placed shall be cleaned thoroughly to the satisfaction of the

engineer. The surface shall be given a uniform application of tack coat in using asphaltic

materials as specified approximately two seconds prior to the placement of the paving

mixture.









2.1.5.6 Hauling equipment

It shall be required that the truck beds be covered while transporting the paving

mixture unless otherwise directed by the engineer. At the discretion of the engineer, the

truck beds may be insulated.

2.1.5.7 Spreading and finishing

The paving mixture shall be delivered at a temperature between 2900 F and 3300 F

unless otherwise directed by the engineer.

The paving mixture shall be dumped directly into the surface treatment paving

machine and spread on the tack coated surface within two seconds of tack coat having

been applied

The paving mixture shall be spread to the depth and width that will provide the

specified compacted thickness, grade and cross section. Placing of the paving mixture

shall be as continuous as possible. The finished surface shall be smooth and of uniform

texture and density.

2.1.5.7 Compaction

Immediately following placement of the paving mixture, the surface shall be rolled

to accomplish a good seating without excessive breakage of the aggregate. A minimum of

three passes with steel wheel rollers is required. The compaction shall be accomplished

prior to the paving mixture cooling below 1800 F. The operation of the rollers shall cause

no displacement or showing of the paving mixture. If the displacement occurs, it shall be

corrected to the satisfaction of the engineer. Sprinkling of the fresh mat shall be required,

when directed by the engineer, to expedite opening the roadway to the traffic. Sprinkling

can be with water or limewater solution.









2.1.5.8 Method of measurement

Paver laid surface treatment, complete in place and accepted will be measured by

the ton. Tack coat will be measured by gallons.

2.2 Porous European Mix (PEM)

PEM also known as porous asphalt is a coarse graded mix with around 4-5 %

binder and 3% of filler. The binder is typically polymer modified. PEM is designed to

contain about 20 % air voids to make it very porous (5,6,7).

The differences between PEM and conventional OGFC are as follows:

* European mixtures generally allow more gap-graded mixtures as compared to the
North American mixtures. However, the GDOT specification for PEM is similar to
the European mixtures.
* All European agencies specify minimum air voids. Air voids in OGFC tend to be
significantly lower than PEM (around 14 %).
* European agencies use modified asphalt binders since modified binders are less
susceptible to draindown, during both construction and service.
* European agencies demand higher standards for aggregate than US agencies. LA
abrasion values are specified from 12-21 %. For OGFC it between 35-40 %
* European agencies specify minimum asphalt content based on durability test
(Cantabro test), which is performed on compacted specimen. A maximum limit is
based on air voids. In contrast to this, for OGFC asphalt content is selected based
on FHWA method. The asphalt content in US mixtures typically varies from upper
five to mid six percent ranges

It has been observed that use of PEM has led to hydroplaning and glare reduction,

noise reduction and increase in skid friction. However, the major disadvantages include

lack of strength and clogging of the pores. The latter problem can be solved using

multiple layers with air voids increasing as we go down.

2.2.1 Advantages of Porous Asphalt

2.2.1.1 Hydroplaning and Glare Reduction

On of the major hazards while driving in rain is aquaplaning. A layer of water

builds between the surface and tire thus causing the vehicle to literally float. As a result,









it becomes difficult to steer and apply the brakes. With porous asphalt it is designed, to

have interconnected air voids for high permeability and higher macrostructure, which

solves the problem of aquaplaning to a great extent.

With porous asphalt, wet weather visibility is significantly improved since there is

little or no free surface water. Usually on unlit roads, the nighttime visibility decreases

due to the water layer on the road surface, which hampers with the retro reflecting

devices. In addition, due to the water logging, mirror reflections are caused by the

headlights of the oncoming vehicles. However, porous asphalt is effective in these

conditions in reducing the glare.

2.2.1.2 Noise Reduction

One of the benefits of porous asphalt mix is the significant noise reduction

compared to the denser mixes due to good absorption potential. Measurements in most

European countries have shown that by using porous asphalt noise levels can be reduced

by approximately 3 dB(A) as compared to the conventional dense asphalt concrete

surfacing (11). These figures apply to passenger car vehicle traveling at 80 Km/hr in dry

conditions. The noise level is influenced by aggregate size, size distribution, permeability

and the condition of the layer. It should be noted that porous asphalt not be used in high

traffic low speed roads since it has been observed that after few years (typically 3 years)

of service all noise reduction benefits are lost because the surface voids get clogged and

the mixture in turn becomes dense graded. However, at high speed, the hydraulic action

flushes the dirt from the pavement voids and reduces clogging. In addition, the cleaning

action would involve spraying or vacuuming.

The vehicle noise is generated by different mechanisms. At high speed, the

vibrations in the tire structure and air pumping in the cavities underneath the tire cause









tire noise. The tread blocks slip on the road surface mostly at the edges of the contact area

where the vertical forces are lower and the tire deforming as the weight is applied or

removed. When the tire rolls forward and the tread blocks snap back to their unloaded

shape it generates vibrations on the tire surface thus generating large noise.

Thus, movement of air in the cavities of tire treads cause air pumping. As the tire

rolls forward the compressed air tries to escape to the sides. As the tread block is lifted

vacuum is created in the tread cavities and the air rushes to fill the voids. If the pavement

texture is smooth, there is less opportunity for the air to leak from underneath the tread

block and the tire noise is accentuated.

2.2.1.3 Skid Friction

Rain may reduce skin friction of road considerably even when no aquaplaning

takes place. Porous asphalt counteracts this effect and even at high speeds the skin

friction with porous asphalt is high on wet roads (1,2).

Finally, on dry roads due porous asphalt gives exceptionally good skidding

properties at higher speeds where macrostructure is very important.

2.2.2 Disadvantages of Porous Asphalt

2.2.2.1 Strength

Porous asphalt does not contribute to the overall structural integrity of the

pavement due to high air voids. In Belgium, it was observed that the moduli of porous

asphalt manufactured with asphalt 80/100, was 73-79 percent of that of wearing course in

conventional asphalt concrete (5).

In Netherlands, the structural behavior of porous asphalt is assessed using the

Department of Public Work's standard multi layer elastic layer design The model

assumes that the fatigue resistance of the road pavements is determined by the lower part









of the structure, which ignores the possibility of fatigue cracks being developed in the

upper part of the structure. Studies have indicated that three aspects need to be

specifically addressed to evaluate the bearing capacity of the porous asphalt, which are as

follows:

2.2.2.2 Initial Stiffness Modulus

Using fatigue test initial E-modulus is determined which is used in the elastic

design model to estimate the contribution of porous asphalt to bearing capacity of the

pavement structure. From the study conducted, it was observed that initial effective

contribution was around 80 to 90 percent of that attainable with gravel asphalt concrete,

depending on thickness of structure. Thus, these mixes tend to have lower stiffness due to

the nature of their structure.

2.2.2.3 Aging and Stripping

As a result of open structure of porous asphalt, the binder is likely to undergo

oxidation to undergo accelerated aging which in turn increases the stiffness of binder

considerably. Also water ingress will lead ton stripping of the lower part of the surface

layer, which affects the cohesive properties as well as the adhesion to the underlying base

course thus impairing the load transfer characteristic of the structure. Based on the studies

conducted it was observed that the weighted effective contribution was 35-40 percent of

that achieved by dense asphalt concrete.

2.2.2.4 Temperature

The suction and pumping action of tires passing over porous asphalt layers, coupled

with wind motion, will promote a continuous circulation of air within pores.

Consequently, the temperature in porous asphalt wearing course is likely to remain closer

to the prevailing air temperature than with the closed surfacing materials. Analysis of









study on temperature gradient on porous asphalt and dense asphalt concrete showed that

the weighted average temperature over a year was found to be about 1VC lower in

pavements surfaced with porous asphalt than inn comparable structures with dense

asphalt concrete wearing course. Thus, the stiffness of porous asphalt is less affected by

warm weather.

Thus the combined of the above three factors implies that porous asphalt can be

expected to contribute to about 50 percent of the equivalent bearing capacity achievable

with dense asphalt concrete.

2.2.2.5 Clogging

The service life of porous asphalt is generally less due to premature clogging of the

voids, which leads to ineffective drainage of the surface water. Also when the surface

pores become plugged, a pavement might fail by asphalt being stripped from the

aggregate surface. However, in Netherlands, they use a new technique of using two layers

of porous asphalt to counter this problem (6). The surface layer uses aggregate that is 4 to

8 mm in size. Directly underneath is another porous asphalt layer containing 11 to 16 mm

sized aggregate. The surface layer allows water and sound to penetrate to larger chamber

of voids in the lower layer. The surface layer has smaller voids to prevent larger materials

from clogging the surface voids. The smaller debris, which enters the surface voids, can

be suctioned out by the hydraulic action of the tires on the pavements.

2.2.3 Performance Related Laboratory Testing

2.2.3.1 Resistance to Plastic Deformation

As a part of laboratory study carried out by University of Cantabria and ESM

Research Centre (8), Spain, the resistance to deformation was found out using wheel-

tracking test at a temperature of 60'C. The percentage of binder used ranged from 4 to 4.5









for the porous asphalt mixes. It was observed that modified asphalt offered greater

resistance to plastic deformation than that of ordinary asphalt. Thus, use of polymeric

asphalt diminishes the effect of post compaction by the traffic, which is sometimes

observed in the porous mixes

2.2.3.2 Resistance to Indirect Tension

The effect of binder on improving resistance to traction was studied in Spain

through IDT testing on Marshall samples as a part of research to study the effect of

special binders on porous mixes (16). The test temperatures chosen were 5 and 45C and

the application rate of 50.8 mm/min. The samples are compacted with energy of 50

Marshall blows per face. The test results showed that the performance of mix made with

polymeric asphalt was better than with ordinary asphalt. The difference was more

prominent at 450C than at 50C.

2.2.3.3 Resistance to Disintegration

Disintegration resistance is tested in the laboratory through the Cantabro test of

wear loss, consisting of testing Marshall samples in the LA abrasion machine and

obtaining the weight loss after 300 drum revolution (9). This test is a standard test for

porous asphalt prevalent throughout Europe. The test is used to predict the maximum and

minimum asphalt content, however the specifications and the test temperature vary with

countries.

2.2.3.4 Adhesiveness

Resistance to stripping is influenced by the adhesiveness between the aggregate and

the binder. For studying resistance to stripping, in Spain and UK, Cantabro test is used to

determine the loss in the test sample that has been submerged in water at a controlled









temperature (around 490C) for 4 days (9). It is observed that modified asphalt has grater

adhesiveness than ordinary asphalt though the losses are higher when immersed in water.

2.2.3.5 Drainage Test

Basket Drainage test (2) has been used in Belgium and Spain for evaluating the

draindown of the porous asphalt mixes. The procedure is as follows:

* The mix is manufactured and compacted in Duriez mold under a pressure of 30
bars
* The molds are then laid on a grid and the set is placed on an oven at 1800C for 71/2
hours, these severe conditions are chosen to simulate occasional cases when the
asphalt is draining through the aggregates
* The asphalt drained through the mix to the grid is recovered and the loss of asphalt
is calculated with respect to the initial binder content.

2.2.4 Mix Design Approach

The design of porous asphalt (6,9,16) is based on

* A minimum binder content to ensure resistance against particle losses and thick
film on the aggregates
* A maximum binder content to avoid binder runoff and have a good drainability in
the mix

Using the maximum abrasion value, a minimum amount of binder is fixed. The

initial selection of the binder type is influenced by the aggregate source and the amount

of binder that needs to be carried. The purpose of using modified binders is to improve

the resistance against particle losses with very open mixtures through higher cohesion

and get a longer durability through thicker films of binder because of higher viscosity. A

reduction in thermal susceptibility is also sought to get higher consistencies with high

temperature and more flexibility wit low temperatures. The layer thickness is typically

around 4 cm (11).

The mix design approach followed by various European countries is as follows:









2.2.4.1 British Design

The mix design approach followed in Britain (6) is as follows:

* Two types of porous asphalt mixes are used; one is 10 mm nominal maximum size
and the other is 20 mm nominal maximum size. The 10 mm is more like that of
other European countries. 20 mm is considerably larger. The gradation is as shown
in Fig.2.2
* The aggregates are specified to provide a hard and durable rock. The LA value for
aggregates must be low, less than 12% or 18%.
* The aggregates must have high polishing resistance and all the aggregates should
have at least two fractured faces.
* Maximum flakiness index of 25 to control aggregate shape
* Both modified and unmodified asphalt binder used. The binder is modified with
Styrene-butadiene-styrene (SBS) or ethylene-vinyl acetate (EVA) modifier
* Specimen compacted using 50 Marshall Blows. Durability is not tested directly but
is controlled through selection of binder grade and minimum asphalt content
* Minimum asphalt content is 4.5 % and minimum air voids is 20%


2.2.4.2 Spanish Design

The mix design approach followed in Spain (6) is as follows:

* There are two gradations of porous asphalt used in Spain. Both are 12.5 mm
nominal maximum size. The P12 has larger gap gradation between 12.5 and 10 mm
sieve. The PA12 has larger gap gradation between 10 and 5 mm sieve. The
gradation is as shown in Fig.2.3
* A maximum LA abrasion value of 20 % for the aggregates
* A maximum flakiness index of 25 %
* Modified asphalt binder used (for high traffic and hot climate 60/70 penetration
asphalt is used. For low traffic and cool temperature, 80/100-penetration asphalt is
used). The modifier used is SBS or EVA
* Specimen compacted using 50 Marshall Blows
* Minimum air voids of 20% selected using Cantabro test with a maximum weight
loss of 25 % at 250C
* Maximum asphalt content is based on the air voids of the compacted specimen.
Typical asphalt content is around 4.5%









2.2.4.3Italian Design

The mix design approach followed in Italy (6) is as follows

* They use just one gradation, a 16 mm nominal maximum size aggregate. All
aggregate are crushed with no natural sand allowed. The gradation is as shown in
Fig.2.4. A maximum LA abrasion of 16% for the aggregates
* A modified binder is used. 6-8 percent is added to 80/100-penetration asphalt. Mix
temperature is 190-200'C. They use SBS as modifier.
* Minimum asphalt content is 4.5 percent based on Cantabro test. The allowable
asphalt content ranges from 4-6 percent. A maximum weight loss of 25 percent is
allowed at 200C. Air voids vary from 18-23 percent
* Moisture damage is evaluated during drainage using Cantabro test. Maximum loss
allowed is 30%

2.2.4.4 Belgium Design

* The mix design consist of first determining the voids in the coarse aggregate and
then measuring with various binder content the voids and the percentage of wear
estimated using Cantabro test
* Aggregates should contain more than 80 percent stones greater than 2mm in
diameter. The proportion of sand (.08 2 mm) should be around 12 percent and
rest should be filler
* A gap grading is to be obtained by omitting 2/7 or 2/10 mm fraction from 0/14 mm
mixture
* Air voids is around 16-18 percent
* Asphalt content is around 4-5 percent (unmodified). For modified binders, it should
be around 5.5 to 6.5 percent.
2.3 Georgia PEM

As already mentioned, OGFC mixes are typically gap-graded and contain high

percentages of single sized coarse aggregate. They typically have high AC content, a

thick AC film, low percentage of material passing the 0.0075 mm sieve and high volume

of air voids (18-22%). The material composition of Georgia PEM is as described below.

2.3.1 Material Selection

2.3.1.1 Composition

The JMF consist of mainly coarse aggregate, typically granite, with small amount

of fines. The binder used should be very stiff such as PG 76-22 made with polymers

(typically SB or SBS). The addition of fibers is desirable since it reduces drain down.















uu

80

60

40 /

20


0.1 1 10 1C
Size (mm)


Size (mm)


Fig. 2.2 British Gradations A) 20 mm B) 10 mm




















Size (mm)


Size (mm)


Fig. 2.3 Spanish Gradation A) P12 B) PA12


rI::::::^ ::7^ -::::e


u/











120

100

80

60

40

20

0


[/


Size (mm)


Fig 2.4 Italian Gradation Band

2.3.1.2 Gradation

The job mix formula is used for the Georgia PEM (21) is as shown in Table 2.5

Table 2.5 Georgia PEM Gradation


Sieve Size Size A 0.45 12.5 mm PEM
(mm) (% Passing)

Max Min
19 3.76 100 100
12.5 3.12 100 80
9.5 2.75 60 35
4.75 2.02 25 10
2.36 1.47 10 5
0.075 0.31 4 1





-

-

-










The 0.45 power chart is as follows


120

100

S80

1 60
a-
40

20


0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
Size ^0.45


Fig. 2.5 Georgia PEM Gradation Band

2.3.1.3 Aggregate Specification

The specifications for basic aggregate properties are as follows (21)

Table 2.6 Aggregate Specifications for Georgia PEM

Parameter Requirement
LA abrasion Loss <50

Soundness Loss (%) <15
Flat ans Elongated <10
Particles (5:1 ratio)
Mica Schist Ratio <10


Note that only silica rich aggregates can be use (e.g. granite). Carbonate- rich

aggregates are excluded. Soundness loss is measured using magnesium sulfate. Typical

loss is less than 2%.

2.3.1.4 Polymer Modified Asphalt cement

GDOT primarily uses two polymers, styrene butadiene (SB) and styrene butadiene

styrene (SBS), to modify cements used in OGFC mixes. The main improvements after









the inclusion of polymers have been 1) increase in binder stiffness by 8-10 times 2) the

softening point of AC has increased by approximately 44oF and 3) the AC is more ductile

and flexible than unmodified AC.

Due to greater viscosity of the polymer blend, temperature requirements in the

design procedure have been increased to 3250F. In addition, a phase angle requirement of

less than 750 has been added to help ensure that polymer modification is used to meet

binder grade requirement. The base asphalt cement is typically modified with 4.0-4.5 %

polymer by weight of AC.

2.3.1.5 Mineral Fibers

Fibers are used in OGFC to stabilize the AC film surrounding aggregate particles in

order to reduce AC draindown during production and placement. GDOT uses mineral

fibers in OGFC at a dosage rate of 0.4 % by weight of total mix.

Thus, the final design requirements have been tabulated in the following table:

Table 2.7 Design requirements for Georgia PEM

Parameter Requirement
Binder Content 5.5 7.0
Polymer SB or SBS (%) 4-4.5
Air Voids (%) 20 24
Drain Down < 0.3
PG 76-22
Fibers (%) 0.2-0.4
Anti Stripping (lime) (%) 1.00


2.3.2 Georgia PEM Mix Design Procedure (GDT 114)

Scope

The design for the Georgia PEM consists of four steps. The first step is to conduct

AASHTO T-245 to determine asphalt cement content then, secondly to determine









optimum asphalt content. The third step to perform GDT-127 and final step is performing

GDT- 56.

Apparatus

1. 13 metal pie pans
2. Oven capable of maintaining 2500 F 3.50 F
3. Oven capable of maintaining 1400 F 3.50 F
4. Beaker glass, 500 ml
5. Glass funnels, top dia = 3.5 in; height = 4.5 in; orifice = 0.5 in with piece of No. 10
sieve positioned at top of funnel neck. Cork stopper to fit the outlet of funnel neck
6. Oil S.A.E No. 10 lubricating
7. Drain Down Equipment as specified in GDT-127
8. Marshall Design equipment as specified in AASHTO T- 245
9. Equipment as specified in GDT 56
10. Balance 5000 gms, 0.1 gm accuracy

Step 1 Surface Capacity (Kc)

* Quarter out 105 g of aggregate representative of material passing through 9.5 mm
sieve and retained on 4.75 mm (No. 4) sieve.
* Dry sample 250'F 3.5F oven to a constant mass and allow to cool
* Weigh out approximately 100 g sample to nearest 0.1 g and place it in a funnel
* Completely immerse sample in S.A.E No. 10 lubricating oil for 5 minutes by
plugging funnel outlet with cork stopper
* Drain sample for 2 minutes
* Place funnel containing sample in 60'C (140'F) oven for 15 minutes of additional
draining
* Pour sample from funnel into a tarred pan, cool, reweigh sample to nearest 0.1 g.
Determine the amount of oil retained as percent of dry aggregate mass
* Use the chart as shown in Fig.6.2. If the apparent specific gravity for the fraction is
greater than 2.70or less than 2.60 apply correction to percentage oil retained using
the formula shown in Fig.2. 1.
* Determine the required asphalt using the following formula
* % AC = 2.0 (Kc)+ 3.5
(No correction applied for viscosity)

Step 2 Modified Marshall Design and Optimum AC

1. Heat the coarse aggregate to 3500F 3.5F, heat the mould to 3000F 3.5 o F and
heat the AC to 330 o F 3.5 o F
2. Mix aggregate with asphalt at three asphalt contents in 0.5 % interval nearest to the
optimum asphalt content establishes in step 1. The three specimens should be
compacted at the nearest 0.5% interval to the optimum and three specimens each at
0.5% above and below the mid interval.









3. After mixing, return to oven if necessary and when 320'F 3.5F compact using 25
blows on each side
4. When compacted, cool to the room temperature before removing from the mold
5. Bulk Specific Gravity:
6. Determine the density of a regular shaped specimen of compacted mix from its dry
mass (in grams) and its volume in cubic centimeters obtained from its dimensions
for height and radius. Convert the density to the bulk specific gravity by dividing
by 0.99707 g/cc, the density of water at 250C
7. Bulk Sp.Gr = W / (7 r2h/ 0.99707) = Weight (gms) x 0.0048417/Height (in)
W = Weight of specimen in grams
R = radius in cm
H = height in cm
8. Calculate percent air voids, VMA and voids filled with asphalt based on aggregate
specific gravity
9. Plot VMA curve versus AC content
10. Select the optimum asphalt content at the lowest point on VMA curve

Step 3 Drain-Down Test

Perform the drain test in accordance with the GDT 127 (Method for determining

Drain Down characteristics in Uncompacted Bituminous Mixtures). A mix with an

optimum AC content as calculated above is placed in a wired basket having 6.4 mm (1/4

inch) mesh openings and heated 14'C (25F) above the normal production temperature

(typically around 350F) for one hour. The amount of cement, which drains from the

basket, is measured. If the sample fails to meet the requirements of maximum draindown

of 0.3%, increase the fiber content by 0.l1% and repeat the test.

Step 4 Boil Test

Perform the boil test according to GDT 56 with complete batch of mix at

optimum asphalt content as determined in step 2 above. If the sample treated with

hydrated lime fails to maintain 95% coating, a sample shall be tested in which 0.5%

liquid anti stripping additive has been used to treat the asphalt cement in addition to the

treatment of aggregate with hydrated lime.















CHAPTER 3
MATERIALS

Five test sections were laid on SR-27 in Highlands County, in July 2003, as part of

friction course study. The section consisted essentially of three kinds of mixes, viz., FC-5

with Limestone, FC-5 with Granite and Novachip. For Georgia PEM, Nova Scotia-

granite was used and the mixes were prepared in the lab. The gradation was exactly same

as specified by the Georgia DOT. The details of the material gradation and binder type

are described in the following sections.

3.1 FC-5 and Novachip

3.1.1 Aggregates

For FC-5, two kinds of aggregates were used: Limestone and Granite. In case of

Novachip, Nova Scotia-granite was used. The gradations for these mixtures are as

shown in Fig 3.1. Clearly the FC-5 Granite and FC-5 limestone seems to have much

coarser gradation that Novachip. The granite was treated with hydrated lime for

prevention against stripping

3.1.2 Binder

For FC-5, the binder used was AC-30 with 12% ground tire rubber. For Novachip

PG76-22, a polymer modified binder, was used. The binder type and binder content used

for the field section is as shown in Table 3.2.The novabond is a polymer modified tack

coat and is same as SBS modified PG 76-22.









Table 3.2 Mixes from the field

Type Binder Type Binder Content
FC-5 Limestone ARB-12 6.40%
FC-5 Limestone ARB-12 6.40%
With
NOVABOND
FC-5 Granite ARB-12 6%
FC-5 Granite ARB-12 6%
With
NOVABOND
NOVACHIP PG 76-22 5%
(SBS
Modified)


3.2 Georgia PEM


3.2.1 Aggregate and Binder

Nova Scotia-granite was used for the mix. Hydrated lime was used as anti stripping

agent. The aggregate gradation used is as shown in Fig 3.1

SBS modified PG 76-22 was used and mineral fiber was added at 0.4% of the mix

as per the requirements. Typically, if the gradation is within the prescribed limit, the

asphalt content is usually around 6% and the air void level is around 20%, which is what

is targeted in these mixes.



















- FC-5
Granite

- FC-5
Limestone

-- Novachip



x--GPEM


Ln On o
CT 0\i)


Sieve Size (raised to 0.45 power) mm


Fig. 3.1 Gradation of FC-5 Granite, FC-5 Limestone, Novachip and Georgia PEM


Lno Co o
1- 1o o
C CO (D
CCC C















CHAPTER 4
DETERMINATION OF COMPACTION LEVEL AND AGING PROCEDURE FOR
FRICTION COURSES

It is a well-known fact that during compaction in the field, a stage is reached where

the aggregate resistance to compaction increases considerably. In other words, there is a

great degree of interlocking between the aggregates. Hence, during compaction in lab it

becomes important to identify the stage at which the mix exhibits this interlocking. This

point of interlocking is called the Locking Point. This concept was first defined by

Vavrik in 1998 for dense graded HMA. The following section focuses on identification of

the locking point for the friction courses.

4.1 Compaction of the Friction Course Mix

Initially each of the FC-5 limestone, FC-5 granite and NOVACHIP mixes were

compacted to 125 gyrations (Ndesign for traffic level 5 as per Superpave criteria) in the

Superpave Gyratory Compactor. This was used to study further the compaction level for

these mixtures.

4.1.1 Compaction Data

The bulk specific gravity of the mixes was obtained from the compaction data.

While calculating Gmb dimensional analysis was done, i.e. Gmbest was used for

calculation of air voids. This was because the mixtures were so coarse that the bulk

specific gravity experiment would not have yielded appropriate value of Gmb. The

volumetric properties of the mixes are as follows:









Table 4.1 Air Voids

Type Gmb Gmm Air
Voids
FC-5 Granite 2.012 2.336 13.87
FC-5 Limestone 1.958 2.441 19.78
Novachip 2.135 2.474 13.7


The graphs of percentage Gmm Vs N for all the pills are as shown in Fig. 4.1 -

Fig.4.3. The compaction curve (% Gmm Vs N) follows logarithmic relationship strongly.

Hence, statistical regression was done using a logarithmic relation. Thus, the equation of

the regression curve was as follows:

%Gmm = m ln(Ngyr) + c

i.e Ngyr = exp((%Gmm-c)/m)

Where m is the slope of the curve at a given gyration and c is a constant.

Now, the locking point (Vavrik, 1998) is defined as the first three gyrations that

are at the same height preceded by two gyrations at same height (the height is in taken

mm up to single decimal place). However based on the above definition, it was found that

compaction of FC-5 with limestone did not yield any locking point which meant that the

limestone had probably been crushed during compaction. Further, even though we could

identify the locking point in case of NOVACHIP and FC-5 with granite, the gyrations

seemed to be on the higher side since the air voids had more or less reached a constant

value by then. (Refer to Table 4.3 for locking points for the mixtures as per Vavrick's

definition). Hence, in order to ascertain whether crushing of the aggregates had taken

place, the gradations of these compacted mixes were determined.











88.00 ---
86.00 R = 0.9968
84.00
82.00
80.00
E 78.00
E 76.00
74.00
72.00
70.00
68.00
66.00
64.00
0 20 40 60 80 100 120 140
Gyrations



Fig. 4.1 Compaction Curve for FC-5 Limestone





y = 3.1847Ln(x) + 70.997
R2 = 0.9944
88.00
86.00
84.00
82.00
80.00
6 78.00
S76.00
g 74.00
72.00
70.00
68.00
66.00
64.00
0 20 40 60 80 100 120 140
Gyrations


Fig. 4.2 Compaction Curve for NOVACHIP











y = 2.6809Ln(x) + 67.651
R2 = 0.9931
88.00
86.00
84.00
82.00
g 80.00 -
E 78.00
^ 76.00
0 74.00
72.00
70.00 -
68.00
66.00
64.00
0 20 40 60 80 100 120 140
Gyrations


Fig. 4.3 Compaction for FC-5 Granite

To check for the breakdown of the aggregates, the gradations of the mixes were

obtained by reflex extraction and compared with the original (ref Fig. 4.7). Looking at the

gradation, it becomes clear that crushing had taken place in case of limestone, though, for

FC-5 with Granite and NOVACHIP there seemed to be no significant difference in the

gradations. However, the compaction curve indicates that for these mixes there is no

significant change in the air voids at higher gyrations. Thus, the locking point should be

way below the existing level of gyrations. Hence, instead of looking at the height of

compaction, we looked at the rate of change of compaction, which was a better indication

of resistance to compaction.

Thus, studying the rate of change of slope of the compaction curve, new locking

point was identified for these friction course mixes. Again, locking point being the point

beyond which the rate of resistance to compaction increases significantly, it implies that

at this stage, the rate of change of compaction decreases significantly and this is what has

been identified in the following study









Table 4.2 Locking point For the Mixtures

Type Locking Point as per Varick( # of
Gyrations)
FC-5 With Lime No Locking Point
FC-5 With 76, 83
granite
NOVACHIP 86, 91, 96 (sample 1)
84, 89, 94 (sample 2)


4.1.2 Initial Study

Since the focus was on resistance to compaction, the exact nature of the compaction

was studied by looking at the rate of change of compaction. Now decrease in the rate of

compaction is directly proportional to the increase in resistance to compaction. This was

essentially used to identify the point of maximum resistance to compaction. It was

observed that the compaction curve became linear beyond a certain gyration. This meant

that compaction had reached a stage where no further decrease in rate of compaction was

possible and this stage was the stage of maximum resistance to compaction. Hence, the

point beyond which the compaction curve became linear was identified and it was

observed that beyond 50-60 gyrations the curve more or less became linear in nature.

This has been presented in Figures 4.4 through 4.6. Thus, the points from visual

identification served as reference values for identifying the locking points for these

mixtures.

It was observed that the compaction curve followed logarithmic trend. To identify

the locking point, the rate of change of slope of compaction curve was used. The stage, at

which the rate of change of compaction was insignificant, was essentially the point of

maximum resistance to compaction.












Point beyond which
curve becomes linear


88.00
86.00
84.00
82.00
80.00
78.00
76.00
74.00
72.00
70.00
68.00
66.00
64.00


0 20 40 60 80 100 120
# of Gyrations



Fig. 4.4 Locking Point for FC-5 with limestone by Visual Observation


Point beyond w which the


88.00
86.00
84.00
82.00
80.00
78.00
76.00
74.00
72.00
70.00
68.00
66.00
64.00


Fig. 4.5 Locking Point for Novachip by Visual Observation


0 20 40 60 80 100 120
# of Gyration










Point Beyond w which curve
becomes linear

88.00
86.00
84.00
82.00 -
E 80.00
E 78.00
( 76.00
0 74.00
72.00
70.00
68.00
66.00
64.00
0 20 40 60 80 100 120 140
# of Gyrations


Fig. 4.6 Locking Point for FC-5 with Granite by Visual Observation

Thus, using the logarithmic regression of the compaction data, the rate of change of

slope can be obtained as,

y = m ln(x) + c

Rate of compaction = dy/dx = m/x (at any x=N)

Rate of change of slope of compaction curve = d2y/dx2 = -m/ X2 (at any x =N)

Based on the above idea the locking point was identified as the point at which two

gyrations at same gradient of slope were preceded by two gyrations at same gradient of

slope. The gradient was taken up to four decimal places (as shown in Table 4.3 for FC-5

Granite). The reason this was chosen as locking point was based on the fact the change in

air voids was insignificant at this stage and that this trend was consistently observed in all

the mixtures. In addition, the compaction level as identified from visual observation was

around 50-60. Thus, based on the above study, the locking points for theses mixtures

were identified as shown in Table 4.4









Table 4.3 Locking Point Based on Gradient of Slope

FC-5 Granite
# of Gyrations Gradient of slope
39 0.0018
40 0.0017
41 0.0016
42 0.0015
43 0.0014
44 0.0014
45 0.0013
46 (LP) 0.0013
47 0.0012
48 0.0012
49 0.0011
50 0.0011


Table 4.4 Locking Points of all Mixtures based on Gradient of Slope

Mixtures Locking Point
FC-5 Limestone 56
FC-5 Granite 46
NOVACHIP 50


Thus based on above concept the locking points for FC-5 with Limestone, FC-5

with Granite and NOVACHIP were 56, 46 and 50 respectively. The specimens were

compacted again to these gyrations and extraction of asphalt was done to observe the

gradations after compaction.

The results of the gradations after extraction are as plotted in Fig. 3.7 (it has been

compared to the original gradations and gradations at 125 gyrations).

For FC-5 Lime even when the gyrations were reduced to 56 from 125, the same

amount of breakdown was observed. This clearly indicated that in case of limestone, the

breakdown occurred in the initial stages itself i.e. at very low gyrations. Hence, even if

the gyrations were to be further reduced, the breakdown was still going to persist. For









FC-5 with granite and NOVACHIP, the gradation looks nearly the same as that of the

original gradation. In addition, the air voids for FC-5 Granite and NOVACHP were

around 22 % and 15 % respectively, which is usually the case for these open graded

mixtures.

Thus, from the above the study it is clear that, though the locking point of each of

these mixtures differed slightly from each other, it was around 50 gyrations. This was

further corroborated by the study done by NCAT on the compaction levels of friction

courses. NCAT suggests 50 gyrations as compaction level for all friction courses

Thus based on our study from visual observation and rate of change of compaction,

NCAT study for friction course, we believe that 50 gyrations should be the compaction

level for friction course mixes.

The steps involved in identifying locking point based on the above-discussed

concept are as follow:

1. Fitting a Logarithmic curve on the compaction curve we get from the Superpave
Gyratory compactor
2. Obtaining the gradient of the slope of the compaction curve by taking the double
derivative of the equation of the regressed curve
3. The locking point is identified as the point at which two gyrations at same gradient
of slope were preceded by two gyrations at same gradient of slope. This is close to
50 gyrations for friction courses

Finally, based on compaction up to 50 gyrations, the air voids obtained for these

mixtures were as follows:

Table 4.5 Air Voids for 50 Gyrations

Type Gmb Gmm Air Voids
FC-5 Limestone 2.012 2.336 17.60
FC-5 Granite 1.958 2.441 21.52
Novachip 2.135 2.474 15.57


















- Original


- 125
Gyrations
--- 46
Gyrations


Sieve Size (raised to 0.45 power) mm


r- U) 0 (0
60dd d -


-*-Original


- 125
Gyrations

-A- 56
Gyrations


)n (n
0)


(N


Sieve Size (raised to 0.45 power) mm


-*-Original


-W-125
Gyrations

---50 Gyrations


Sieve Size (raised to 0.45 power) mm


Fig. 4.7 Gradations after Extraction for A) FC-5 Granite B) FC-5 Limestone C) Novachip


100

Z 80

C- 60

40

20

a 0


D) 100
-
*I 80
-
D 60
-
40
w 2-
S20

- 0
0


0) 100

80

60
0)
40

20

a 0









4.2 Asphalt Content Determination

The asphalt contents for FC-5 with limestone, Granite and NOVACHIP were 6%,

6.4 % and 5% respectively. For the friction courses, when laid with tack coat, the

interface between the fiction course and the underlying layer becomes nearly saturated.

The tack coat is heavily polymer modified and essentially contains PG 76-22 binder

modified with SBS. Hence, to replicate this field condition in the lab it was essential to

nearly saturate the existing mixes with PG 76-22. This tack coat is primarily used in

NOVACHIP, but for the field project, the very same tack coat was used even for FC-5

with limestone and Granite.

The volumetric calculation for determining weight of asphalt to be added was

based on maintaining same VMA and reaching the target air void level. The derivation as

shown below:

Gmml = Initial theoretical maximum specific gravity

Gmb 1= initial bulk specific gravity of the mixes

Gmb2= bulk specific gravity after addition of Pg76-22

W1= Total weight of mix before addition of PG76-22

Vt = Total volume of mix

Vl= volume of aggregate and asphalt before addition of PG76-22

Vairl = volume of air voids before asphalt addition

Vair2 = volume of air voids finally after adding PG76-22

a2 = Final Percentage of Air Voids Desired

6Was = weight of PG76-22 added

Thus Gmbl= W1/(V1+ Vairl)









V1+ Vairl = volume of compacted specimen =Vt

Vt = W /Gmb 1..... .......... .... .. ........ ........ ........... .. (1)

Vl= W1/Gmml

Now again in this mix we add additional PG 76-22. The total volume Vt remains

the same but the percentage distribution of air voids and asphalt content changes. Thus,

the Gmm of the mix changes

Thus Gmm2 = final maximum theoretical specific gravity

= (W1+6Was)/(V1+ 6Vas)

Now Gmb2 = (1-a2/100) *Gmm2

= (1-a2/100) (W1+6Was)/(V1+ 6Vas)..................................(2)

Where 6Vas = volume of asphalt added = 6Was/Gb

And Gb= specific gravity of PG76-22 = 1.028

Also Gmb2 = (W1+6Was)/Vt (the final volume is same)...........................(3)

Thus equating (2) and (3) and substituting (1) we can get the amount of asphalt to

be added

6Was = ((1-a2/100)*Vt-V1)*Gb

4.3 Mixing and Compaction

As already mentioned, to replicate the interface consisting of tack coat, it was

decided that we add PG 76-22 since the tack coat contained the same binder. Hence was

decided that the air voids in the saturated samples be reduced to 2 %. The mixing and

compaction temperature for all the three mixes was 3200F.









During mixing, it was observed that amount of drain down was excessively high

and it was difficult to work with these mixes at such high asphalt content. Further,

compaction was virtually impossible because of such high drain down.

Hence, it to was decided that the asphalt added should be such that air voids were

reduced by 50% in case of Novachip and 25 % in case of FC-5 with granite and

limestone. Again, initially the mixing and compaction temperature was kept as 3200F.

However, the amount of drain was still significant. Then it was decided that both mixing

and compaction temperature be brought down to 2550F. However, because of polymer-

modified asphalt, mixing became very difficult since the asphalt was very viscous at such

a low mixing temperature. Finally, a mixing temperature of 3200F and compaction

temperature of 2550F was adopted and workability of these mixtures improved

significantly. However, since the asphalt content was still very high, the loss in asphalt

due to drain down was in the range of 10-15 grams.

It was observed that after addition of asphalt these mixes, especially FC-5 with

Granite, they became very sensitive to the compaction temperature. For the mixes

compacted without addition of asphalt, number of gyrations was determined as 50.

However, for the saturated samples, the gyration level varied from 60-90. This indicates

that the rate of reduction of temperature in case of polymer-modified asphalt is very high.

Finally, after compaction the sample was not immediately retrieved from the

mould. It was allowed to cool down for around 1 hr 15 min before ejecting the sample out

(for unsaturated field mix the pills were allowed to cool down for 45 minutes before

retrieving them). This was because it was observed that, because of high air voids, these

mixes collapsed when they were retrieved from the mould immediately after compaction









The final air void level and Gmm for the mixtures after compaction was as follows

Table 4.7 Weight of PG 76-22 to be added

Type Initial Final


Weight of Gmb Gmm Air Air %Gmm Gmm Asphalt Compaction
Mix Voids Voids Added Height
(gms) (0o%) (%%)) N(gms) (mm)
FC-5 4800 1.923 2.336 17.68 13.26 86.740 2.269 113.50 141.250
Limestone
FC-5 4700 1.916 2.441 21.51 16.13 83.869 2.350 135.59 138.813
Granite
Novachip 4800 2.089 2.474 15.56 7.78 92.219 2.352 183.79 130.000


4.4 Long Term Oven Aging of Friction Course

The mixes were taken from the wood way and hence they had already undergone

short-term aging. The mixes were subjected to long-term aging according to LTOA-

AASHTO PP2-94. However, the mixes being very course and open, there was a

possibility of these mixes falling apart during aging. Hence, a procedure was developed

to contain the compacted pills from falling apart during aging.

A wired mesh with opening of 1/8" and steel clamp was used. The mesh size was

chosen in order to ensure that there is good circulation of air within the sample for

oxidation and at the same time, to prevent the smaller aggregate particles from falling off

from the mesh. The specimen was rolled over twice by the mesh and two clamps were

used to contain the specimen without applying excessive pressure on it. The whole

system looked as shown in Figures 4.8-4.9






























Fig. 4.8 1/8 Mesh used for containing the Pill


Fig. 4.9 Pill contained with Mesh
















CHAPTER 5
INDIRECT TENSILE TESTING

The fracture evaluation of the friction courses was done within the framework of

HMA fracture mechanics (16). The HMA fracture mechanics deals with the concept of

threshold, which is defined as material's state between micro-damage and macro-crack

development and is dealt in terms of energy. There are two energy limits, which define

failure, viz., Dissipated Creep Strain Energy (DSCE) and Fracture Energy (FE). DSCE is

chosen as criterion under repeated loading condition while FE is selected under critical

loading condition as shown in Figure 4.1


CASE 1
Repeated Load
Cyclic Fatigue




Fail

. - - -


CASE 2
Fail at the
Critical Load




Fail

- - - - -


CASE 3





---------------------- FEthreshold

No Failure

_________-__________-_ DEthreshold


N (Number of Load Replications)


Fig. 5.1 Effect of Loading Condition


Energy









The key points of the HMA Fracture Mechanics Model can be summarized as

follows:

* If the threshold as defined in this model is not exceeded, micro damage is healable,
once the threshold is exceeded, macro damage s not healable
* Under repeated loading conditions, DSCE can be used as threshold and it can be
easily obtained from strength tests using the Superpave Indirect tensile test System
(IDT)
* Asphalt being a viscoelastic material, crack imitation and propagation cannot be
distinguished, cracks grow discontinuously (i.e. crack grows in step wise manner)
* Under critical loading condition, FE obtained from strength test can be used as
threshold
* This model handles realistic loading condition and healing effects on asphalt
pavements using the DSCE and FE as criterion
* All parameters needed to describe crack growth are obtained from relatively simple
Indirect Tensile test (IDT) (i.e. resilient modulus, creep response- m-value, fracture
energy to failure and tensile strength)

5.1 Sample Preparation

The friction course being very porous, in order to avoid end effects, it was decided

that the sample thickness be around 1.5 -2 in. However, in case of FC-5 with limestone

with additional asphalt, the thickness was reduced to 1.0 in because it was observed that

due to high asphalt content, the limestone stone was being crushed at top before failure.

Therefore, the thickness was reduced which then solved this problem.

A cutting device, which has a cutting saw and a special attachment to hold the

pills (Figure 5.2), was used to slice the pill into specimens of desired thickness. Two two-

inch samples were from each specimen. Because the saw uses water to keep the blade

wet, the specimens were dried for one day at room temperature to achieve the natural

moisture content. Before testing, Specimens were placed in the humidity chamber for at

least two days to negate moisture effects in testing.

Gage points were attached to the samples using a steel template and vacuum

pump setup and a strong adhesive (Figure 5.3). Four gage points were placed on each









side of the specimens at distance of 19 mm (0.75 in.) from the center, along the vertical

and horizontal axes. A steel plate that fits over the attached gage points was used to mark

the loading axis with a marker. This helped placing the sample in the testing chamber

assuring proper loading of the specimen (Figure 5.4).


Fig. 5.2 Cutting Device



































Fig. 5.3 Gauge Point Attachment



































Fig. 5.4 Marking Loading Axes


5.2 Test Procedures

Standard Superpave IDT tests (13,14) were performed on all mixtures to determine

resilient modulus, creep compliance, m-value, Di, tensile strength, failure strain, fracture

energy, and dissipated creep strain energy to failure. The tests were performed at 100C.

5.2.1 Resilient Modulus Test

The resilient modulus is defined as the ratio of the applied stress to the recoverable

strain when repeated loads are applied. The test was conducted according to the system

developed by Roque et al (13) to determine the resilient modulus and the Poisson's ratio.

The resilient modulus test was performed in load control mode by applying a repeated

haversine waveform load to the specimen for a 0.1 second followed by a rest period of

0.9 seconds. The load was selected to keep the horizontal strain in the linear viscoelastic









range, in which horizontal strain is typically 150 to 350 micro-strains. The procedures for

resilient modulus test are as follows:

1. The specimens compacted are cut parallel to the top and bottom faces using a
water-cooled masonry saw to produce 2 inches thick specimens having smooth and
parallel faces.
2. Four aluminum gage points are affixed with epoxy to each trimmed smooth face of
the specimen.
3. Test samples are stored in a humidity chamber at a constant relative humidity of 60
percent for at least 2 days. In addition, specimens are cooled at the test temperature
for at least 3 hours before testing.
4. Strain gauges are mounted and centered on the specimen to the gage points for the
measurement of the horizontal and vertical deformations.
5. A constant pre-loading of approximately 10 pounds is applied to the test specimens
to ensure proper contact with the loading heads before test loads are applied. The
specimen is then tested by applying a repeated haversine waveform load for five
seconds to obtain horizontal strain between 150 to 300 micro-strains. If the
horizontal strains are higher than 050 micro-strains, the load is immediately
removed form the specimen, and specimen is allowed to recover for a minimum
minutes before reloading at different loading level.
6. When the applied load is determined, data acquisition program begins recording
test data. Data are acquired at a rate of 150 points per seconds.
7. The resilient modulus and Poisson's ratio are calculated by the following equations,
which were developed based on three dimensional finite element analysis by Roque
and Buttlar (1992). The equation is involved in the Superpave Indirect Tensile Test
at Low Temperatures (ITLT) program, which was developed by Roque et al (1997).

Px GL
MR =
AH x t x D x C,,mn


Where,

MR = Resilient modulus

P = Maximum load

GL = Gauge Length

AH = Horizontal Deformation

t, D = Thickness, Diameter


Ccomp= 0.6354x (X/Y) 1-0.332









5.2.2 Creep Test

Creep compliance is a function of time-dependent strain over stress. The creep

compliance curve was originally developed to predict thermally induced stress in asphalt

pavement. However, because it represents the time-dependent behavior of asphalt

mixture, it can be used to evaluate the rate of damage accumulation of asphalt mixture.

As shown in Figure 5.5, DO, Dl, and m-value are mixture parameters obtained from

creep compliance tests. Although Dl and m-value are related to each other, Dl is more

related to the initial portion of the creep compliance curve, while m-value is more related

to the longer-term portion of the creep compliance curve.

The m-value has known to be related to the rate of damage accumulation and the

fracture resistance of asphalt mixtures. In other words, the lower the m-value, the lower

the rate of damage accumulation. However, mixtures with higher m-value typically have

higher DCSE limits. The creep compliance is a time dependant strain, E(t), divided by a

constant stress. That is, the inverse of the creep compliance, which is called creep

stiffness, is a kind of stiffness. According to the analysis conducted by Roque et al (13),

MR is higher than creep compliance stiffness at 1 second.

The Superpave Indirect Tensile Test at Low Temperatures (ITLT) computer

program was used to determine creep properties of the mixtures. The test was conducted

in a load control mode by applying a static load. The load was selected to keep the

horizontal strain in the linear viscoelastic range, which is below a horizontal strain of 500

micro strains.

The test procedure was presented by Roque et al (1997). The procedures for

indirect tensile creep test consist of the following steps:









* The preparation of test samples and the pre-loading are same as those for resilient
modulus test
* Apply a static load for 1000 seconds. If the horizontal deformation is greater than
180 micro inch at 100 seconds, the load is immediately removed from the
specimen, and specimen is allowed to recover for a minimum 3 minutes before
reloading at a different level. At 100 sec, the horizontal deformation should be less
than 750 micro inches
* When the applied load is determined, the data acquisition program records the
loads and deflections at a rate of 10 Hz for the first 10 seconds, 1Hz for the next
290 seconds, and 0.2 Hz for the remaining 700 seconds of the creep test.
* The computer program, ITLT, was used to analyze the load and deflection data to
calculate the creep compliance properties. Creep compliance and Poisson's ratio
are computed by the following equations.


AH x t x D x Ccomp
D (t) =
Px GL



v = -0.1+1.480 x (X/Y)2- 0.778 x (t/D)2 x (X/Y)2

Where, D (t)= Creep Compliance

5.2.3 Strength test

Failure limits such as tensile strength, failure strain, and fracture energy were

determined from strength tests using the Superpave IDT. These properties are used for

estimating the cracking resistance of the asphalt mixtures. The strength test was

conducted in a displacement control mode by applying a constant rate of displacement of

50 mm/min for field mix and 100 mm/min for saturated mix until the specimens failed.

The horizontal and vertical deformation and the applied load are recorded at the rate of

20 Hz during the test.

The maximum tensile strength is calculated as the following equation.



2 x P x Csx
St =
rE xbxd














D(t) =Do + Dt1T


D0+D, L


Time (t)


LoDg DIr


Log (D-Di)


0 Log (time)

where:
D(t) = Creep compliance at time, t
Do., DI, m = Power model constants


Fig. 5.5 Power Model for Creep Compliance


D(t)









Where,

St = Maximum Indirect tensile Strength

P = Failure load at first crack

Csx, = 0.948 0.01114 x (b/D)- 0.2693 x v + 1.436 x (b/D)v

b, D = Thickness, diameter

From the strength test and the resilient modulus test, fracture energy and dissipated

creep strain energy can be determined. Fracture energy is a total energy applied to the

specimen until the specimen fractures. Dissipated creep strain energy (DCSE) is the

absorbed energy that damages the specimen, and dissipated creep strain energy to failure

is the absorbed energy to fracture (DCSEf). As shown in the Figure 3.9, fracture energy

and DCSEf can be determined as described below. The ITLT program also calculates

fracture energy automatically.


St MR X Ef St
MR= > o =
Cf En MR


Elastic Energy (EE) = (1/2) x St x (Ef go)


Fracture energy (FE) = S (e) de
0

Dissipated Creep Strain energy (DSCE) = FE EE

Where, St = Tensile Strength

gf = Failure Tensile Strain










Fracture Energy = DCSEf+ E.E.

St










Dissipated Creep Strain Energy Elastic
to failure Energy



r- ef

Fig.5.6 FE and DCSE from Strength Test

5.3 Issues Related with IDT Testing of Friction Course

As already mentioned, the specimen thickness were kept between 1.5 2 in for the

all field mixes and for saturated FC-5 Granite and Novachip to avoid end effects. The

reason, this thickness range was chosen was because these graded friction courses are so

open that at the point of contact with the loading strip, the surface might be so weak that

local failure might occur under stress concentration. Thus, there was a possibility of not

getting true tensile failure. However, for saturated FC-5 limestone, the thickness was kept

close to one inch because with 2-inch sample the aggregate at the top was being crushed.

This was because the mix had such high asphalt content that, at 100C, the polymer-

modified asphalt became stiff and before the failure could propagate through the mastic,

the lime rock at the top was crushed. Hence, the thickness was reduced to lower the peak






60


load to failure and avoid end effects. Another reason for the end effect failure in

limestone is the loading rate, which has been discussed in the following paragraph.

Now for the field mix, the loading rate in strength test was 50-mm/ min. However,

for saturated sample, the samples were failing due to end effects at that loading rate. This

was observed primarily because of shear failure rather than true tensile failure. Hence, the

loading rate was increased to 100 mm/min, and end effect failure was no longer observed.














CHAPTER 6
MIX DESIGN APPROACH FOR GEORGIA PEM

The Georgia DOT uses Marshall's blow for design of PEM. However, for the

Georgia PEM mix design, we used Superpave Gyratory Compactor. Further, the surface

capacity determination, as explained in literature review, is not needed if the gradation is

within the desired gradation band. The over all mix design approach is explained in the

following chapter.

6.1 Evaluation of compaction Level for Georgia PEM

The compaction level used for Georgia PEM was 50 gyrations. Now, as already

stated, the approach used in determination of compaction level was based on rate of

change of the slope of compaction curve. The Locking Point was identified as the point at

which two gyrations at same gradient of slope were preceded by two gyrations at same

gradient of slope. Thus based on above definition, the locking point for the three pills of

Georgia PEM was identified as given in Table 6.1

Thus, from the table its clear that the compaction level for Georgia PEM is also

around 50 gyrations, which was used for the design.

6.2 Mix Design Approach

Scope

The design for the Georgia PEM consists of three steps. The first step is to conduct

AASHTO T-245 to determine asphalt cement content then, secondly to determine

optimum asphalt content. The final step is performing GDT-127









Table 6.1 Locking Point for Georgia PEM

Sample No. of Gradient of the
Gyrations Slope
1 43 0.0015
44 0.0015
45 0.0014
146 0.0014
2 44 0.0014
45 0.0014
46 0.0013
47 0.0013
3 47 0.0011
48 0.0011
49 0.0010
50 0.0010


Initial Asphalt Content

Based on number of experiments, the GDOT suggests using 6 % as the first

estimate of asphalt content. It has been observed that if the gradation is within the

specified limits, the initial estimate is comes out to be 6%.

Modified Marshall Design and Optimum AC:

1. Heat the coarse aggregate, the mould to and the AC to 330 oF 3.5 F
2. Mix aggregate with asphalt at three asphalt contents, viz., 5.5%, 6% and 6.5%. Just
before mixing, add the required amount of mineral fibers to the aggregate. Prepare
three samples at each of the asphalt content
3. After mixing, return to oven for two hours for STOA at 320 o F 3.5 o F Then
compact using the Superpave Gyratory Compactor 50 gyrations
4. When compacted, cool to the room temperature before removing from the mold. It
typically takes 1 hour 45 min to cool down.
5. Bulk Specific Gravity:
6. Determine the density of a regular shaped specimen of compacted mix from its dry
mass (in grams) and its volume in cubic centimeters obtained from its dimensions
for height and radius. Convert the density to the bulk specific gravity by dividing
by 0.99707 g/cc, the density of water at 25 'C
7. Bulk Sp.Gr = W / (n r2h/ 0.99707) = Weight (gms) x 0.0048417/Height (in)
W = Weight of specimen in grams
8. R= radius in cm
9. H = height in cm
10. Calculate percent air voids, VMA and voids filled with asphalt based on aggregate
specific gravity









11. Plot VMA curve versus AC content
12. Select the optimum asphalt content at the lowest point on VMA curve

Drain-Down Test

Perform the drain test in accordance with the GDT 127 (Method for determining

Drain Down characteristics in Uncompacted Bituminous Mixtures). A mix with an

optimum AC content as calculated above is placed in a wired basket having 6.4 mm (1/4

inch) mesh openings and heated 14 'C (25 'F) above the normal production temperature

(typically around 350OF) for one hour. The amount of cement, which drains from the

basket, is measured. If the sample fails to meet the requirements of maximum draindown

of 0.3 %, increase the fiber content by 0.1 % and repeat the test.

Thus based on the above design procedure, the optimum asphalt content was

determined to be 6%. The fiber content was 0.4% of the total mix. The drain down was

minimal (around .01%). The final mix design in as shown in Fig. 6.1

It is recommended by GDOT that the asphalt content should not be below 6%

because of coating issues. The film thickness at 6% asphalt was around 39.6 micron. The

drain down was .01%, which is well below the limit. Thus based on 6% AC, pills were

prepared for IDT testing.

























*iI A _-"o _AC GCmiii Gmh_ MAA 'TNI VFA
2.641 5.5 2.442 1.936 30.74 20.72 32.60
6 2.414 1.961 30.23 18.78 37.86
6.5 2.389 1.967 30.38 17.68 41.82


VMA


VTM


VFA


3080
3070
30 60
< 3050
> 3040
3030
3020
30 10
54 56 58 6 62 64 66
%AC


21 00


20 00


i 1900


1800


1700
54 56 58 6 62 64 66
%AC


4400
42 00
4000
S3800
> 3600
3400
32 00
30 00
5 55 6 65 7
%AC


Optimum AC



Mineral Fiber


6%


0.4%
of Total Mix


Mixing Temperature



Compaction Temperature


330 F



325 F


Fig 6.1 Final Mix Design for Georgia PEM














CHAPTER 7
FINDINGS AND ANALYSIS

The results obtained from the IDT test have been analyzed using ITLT software

developed at University of Florida. Using the results from the software DCSE and Energy

Ratio were calculated. Energy Ratio is dimensionless parameter that serves as single

criteria for cracking performance of mixtures in pavements. It is defined as follows:

ER= DCSEf/ DCSEmin

Where, DCSEf is the dissipated creep strain energy threshold, and DCSEmin is the

minimum dissipated creep strain energy, which is given by:

DCSEmin = m2.98 Di/ A,

Where, m and D1 are creep compliance parameters.

A is a parameter dependent on tensile stress and tensile strength, St

A = 0.0299 5 -3.10 (6.36-St) + 2.46 10-8

Where, a is the applied tensile stress.

The applied tensile stress is taken at the bottom of the AC layer and is very much

dependent on the stiffness of the AC layer. In case of friction course, the whole pavement

structure can be regarded as a composite consisting of friction course and AC layer

beneath it. Thus, the stress at the bottom of the AC layer should be considered for ER

calculation for the friction course. It should be noted that the ER was determined for

dense graded mixtures only and it has not been truly calibrated for friction courses.

Thus, a typical pavement structure with the given loading condition was considered

for determining ER for all the mixes. This has been illustrated in Fig7. 1.










Pressure = 120 psi,
Load = 9000 Ilbs




Friction Course. Mr.v (from IDT) 0.75 in


AC Layer, E = 800000 psi 8 in
v = 0.4



Base course E = 500000 12 in
v = 0.35



Subgrade, E = 10000 psi


Fig 7.1 Typical Pavement Structure for Stress Calculation

The thickness for FC-5 and Novachip was taken as 0.75 in and for Georgia PEM it

was taken as 1.25 in. The stresses were calculated using KENLAYER and the results

have been attached in Appendix C.

The other parameters used for evaluation are the FE, DCSE, tensile strength, failure

strain and creep strain rate. The creep strain rate for a 1000 sec creep test is calculated as

follows:

ds(t)/dt/o = dD(t)/dt = D1 m (1000)m-1

Where, ds(t)/dt/o = strain rate per unit stress

D(t) = creep compliance

7.1 Evaluation of FC-5 and Novachip Field Mix

The test results for the friction courses are presented in Figures 7.1 through 7.7.

The three field mixes were also subjected to long-term oven aging to study the effects of

aging on their fracture resistance. In addition, to replicate the interface between the









surface mixture and the underlying HMA, which contains the polymer, modified tack

coat, additional PG76-22 was added to the field mix to reach the desired target air void

level. The results for the saturated mixes are presented in Figures 7.15 through 7.21

7.1.1 Unaged Field Mix

As shown in Fig 7.4, for FC-5 Granite and FC-5 Limestone, the ER is around 1.5

which indicates good field performance. For Novachip, the ER is close to 3, which

indicates that it is a better performing surface course than FC-5. In case of FC-5 Granite,

the failure strain is around 3000 micro strain and DCSE is close to 3 KJ/m3. For

limestone, the DCSE is very low (just below 1 KJ/m3). This is primarily because of low

failure strain (around 950 micro strain). For Novachip the DCSE is around 4.5 KJ/m3

with a high failure strain of 3200 micro strain. As far as the tensile strength and resilient

modulus are concerned, Novachip has higher values as compared to FC-5 granite and

limestone.

Finally, Novachip and FC-5 Granite have significantly high creep rate (close to 3e-

08/psi-sec) as compared to FC-5 Limestone. For Granite, this is one of the reasons for

low ER value. For Novachip high creep rate is compensated by high DCSE, which in turn

is reflected in its high ER value.

7.1.2 Aged Field Mix

In case of granite, there is substantial decrease in ER due to long-term aging. As

shown in Fig 7.5, for granite-aged mix, the failure strain reduced by half as compared to

unaged mix. As a result, the DCSE also plummeted to 1 KJ/m3 which resulted in really

low ER value of 0.7.











5.00
4.50
4.00
3.50
C 3.00
2 Unaged
2.50
0 Aged
W 2.00
1.50
1.00
0.50
0.00
FC-5 G FC-5 L NOVACHIP




Fig 7.2 Fracture Energy for Field Mix





5.00
4.50
4.00
3.50
E
S3.00
S25 Unaged
2.50
SOAged
W 2.00
S1.50
1.00
0.50
0.00
FC-5 G FC-5 L NOVACHIP


Fig 7.3 Dissipated Creep Strain Energy for Field Mix











4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00


Fig 7.4 Energy Ratios for Field Mix


FC-5 G FC-5 L


SUnaged
0Aged


NOVACHIP


Fig 7.5 Failure Strain for Field Mix


FC-5 G


FC-5 L


4000.0

3500.0

3000.0

2500.0

2000.0

1500.0

1000.0

500.0

0.0


* Unaged
O Aged


NOVACHIP











2.50


- 2.00
CL
a
I
S 1.50


1.00


- 0.50


0.00


FC-5 G


FC-5 L


Fig 7.6 Tensile Strength for Field Mix


10.00
9.00
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00


FC-5 G


FC-5 L


Fig 7.7 Resilient Modulus for Field Mix


* Unaged
O Aged


NOVACHIP


* Unaged
O Aged


NOVACHIP


I











5.00E-08
4.50E-08
a 4.00E-08
a. 3.50E-08
W 3.00E-08
0) EUnaged
2.50E-08 -
I 0 Aged
S2.00E-08 -
S1.50E-08
S1.00E-08


0.00E+00
FC-5 G FC-5 L NOVACHIP



Fig 7.8 Creep Rate for Field Mix

The creep rate however increases with aging for granite. This could be attributed to

the fact there are critical points in the matrix of aggregate and mastic, wherein aging

takes place locally. These critical points are points of stone-stone contact. Further, there

are regions in the matrix where aging is not so effective due to large globule of mastic.

Hence, due to aging, we have a structure with varied stiffness and hence different regions

would respond differently to load application. Thus, any kind of damage to these critical

points, which govern the creep behavior of the specimens, causes damage in the

specimens, resulting in higher creep rate. This is illustrated in Fig 7.9 (a).

For Novachip, aging seems to have little effect on FE, DCSE and ER. However,

interestingly, for limestone, aging increases ER by factor of two (ER for aged limestone

is around 4.2). This is primarily due to higher failure strain, DCSE, and lower creep rate.

To explain this phenomenon, a study was done on the effect of aging on these mixes,

which is presented below.









The asphalt content of the friction course is around 6 %, which is high as compared

to any dense graded mixture. Further, limestone in general has a very rough texture and

has lot of crevices and pores in it. Now, the long-term oven aging temperature is around

85 C and at such a high temperature, the asphalt flows, occupies the crevices and the

pores in the aggregate. This is illustrated in Fig 7.9 (b). Thus, essentially, with aging in

laboratory, the absorption by limestone increases and the aggregates become more

ductile. In addition, the bonding between the aggregates also increases. Now, visual

observation of the cracked specimens indicated that the failure happened through the

aggregate. Thus, because of increased absorption due to aging, limestone becomes more

ductile and hence the failure strain increases resulting in increased DCSE and ER.

It should however be noted that in the field, the temperature varies from anywhere

between 10-60 C and only in summers the pavement temperature approaches the higher

end. In addition, these mixtures being surface mixtures will certainly not be subjected to

very high temperatures. Thus, we believe that it is only in laboratory aging, such results

can be observed and that it is not a realistic representation of field condition

To corroborate the laboratory results further, another set of testing was done on

limestone under the following conditions i.e. 1) Unaged, 2) 2-Day aging, 3) 5-Day aging

(LTOA) 4) 10-Day aging. The results are presented in Figures 7.10 through 7.16.

The results for unaged and long-term oven aged specimens showed the same trend

as before. Clearly, looking at the ER, we see that 2-Day and 5-Day aged mix have higher

ER as compared to the unaged mix. The same is the case with DCSE and failure strain.

However, for the 10-day aged mix, the DCSE falls below that of unaged mix, though the

ER remains higher.












Critical Point where
maximum aging will
occur


Area where aging
might not be effective




Mastic




(A)








Absorbed
asphalt in
Crevices





(B)

Fig 7.9 Effect of aging on A) FC-5 Granite B) FC-5 Limestone









The 2-day aged mix has the highest ER and then it decreases with aging. The same

trend is observed for DCSE and failure strain. Thus, as already stated before, during the

process of laboratory aging, the asphalt starts flowing and occupies the crevices and pores

in the lime rock making the mix more ductile. However, as aging progresses, the

absorbed asphalt also starts getting aged and we see decrease in the failure strain, and

hence the DCSE, for the 5-day and 10-day aged mix. This again is reflected in the lower

ER values. The resilient modulus increases slightly with aging. The tensile strength

increases for the aged mixes. For this new set of testing, the creep rate of the unaged mix

seems to be on the higher side as compared to field mix. However, the 5-day aged mix

seems to have around the same creep rate as obtained before for aged field mix. The 10-

day aged mix has the lowest creep rate.

7.1.3 Saturated Unaged Mix

As already mentioned, modified PG76-22 was added to the mixes to see the effect

of Novabond on the fracture resistance of the mixes. Clearly, for FC-5 granite and

Novachip, Novabond increases the FE, DCSE and failure strain (Fig 7.18-7.24). Infact,

for Novachip, the FE increases to 9 KJ/m3, which is twice as much as for unaged field

mix. The ER for granite (around 2) does not increase substantially, possibly because of

increase in creep rate with the addition of asphalt (Fig 7.17). However, for Novachip, the

creep rate remains more or less same as that of the field mix and DCSE increases

significantly to 9 KJ/m3 hence we see a huge jump in the ER value (slightly above 6). For

limestone, the failure strain increases to around 1400 micro strain and DCSE increases

slightly to around 1.2 KJ/m3. The ER for limestone saturated however decreases as

compared to the field mix. This is primarily because of the high creep rate (around 3e-










08/psi-sec) due to addition of asphalt. It should be noted that for the saturated mixes, the

resilient modulus decreases due to increased asphalt content.

7.1.4 Saturated Aged Mix

As in case of field mix, aging decreases ER for Granite (slightly above 1) and

Novachip. This is because of significant decrease in the failure strain and hence in the

DCSE. In case of limestone, the aged mix has higher ER than the unaged, as was the case

with field mix. However, the aged limestone mix has lower DCSE and failure strain than

unaged mix. The reason for higher ER is because of high creep rate of unaged saturated

mix. The creep rate of aged saturated limestone mix was same as the field mix. Another

interesting thing to note is that, the FE, DCSE and ER of aged field limestone mix are

much greater than saturated limestone mix. A possible explanation for this is that, at high

asphalt content, the limestone loses the benefit of its rough surface texture and the mix

with aging becomes very brittle because of high asphalt content. This results in lower

failure strains and hence lower fracture energy.


2.00
1.80
1.60
1.40 -
E 1.20 -
1.00 -
w. 0.80 -
0.60 -
0.40 -
0.20 -
0.00
Unaged 2 Day Aging 5 Day Aging 10 Day Aging


Fig. 7.10 FE for FC-5 Lime at Various Stages of Aging











2.00
1.80
1.60
1.40
E 1.20
1.00
C 0.80
0.60
0.40
0.20
0.00
Unaged 2 Day Aging 5 Day Aging 10 Day Aging



Fig. 7.11 DCSE for FC-5 Lime at Various Stages of Aging


1600.0

1400.0

1200.0

1000.0

S 800.0

600.0

400.0

200.0

0.0


Unaged 2 Day Aging 5 Day Aging 10 Day Aging


Fig. 7.12 Failure Strain for FC-5 Lime at Various Stages of Aging







77



4.00

3.50

3.00

2.50

> 2.00

1.50

1.00

0.50

0.00
Unaged 2 Day Aging 5 Day Aging 10 Day Aging



Fig. 7.13 Energy Ratio for FC-5 Lime at Various Stages of Aging


2.00
1.80

1.60
1.40 -

S1.20 -
S1.00 -
0.80 -

o 0.60 -
a,
0.40

0.20 -

0.00
Unaged 2 Day Aging 5 Day Aging 10 Day Aging



Fig. 7.14 Tensile Strength for FC-5 Lime at Various Stages of Aging












10.00
9.00
8.00
7.0-
7.00
5.0-
6.00

3.0-
5.00
4.00
3.00
2.00
1.00
0.00


Unaged 2 Day Aging 5 Day Aging 10 Day Aging


Fig. 7.15 Resilient Modulus for FC-5 Lime at Various Stages of Aging


1.40E-08

1.20E-08

1.00E-08

8.00E-09

6.00E-09

4.00E-09

2.00E-09

0.OOE+00


Unaged 2 Day Aging 5 Day Aging 10 Day Aging


Fig. 7.16 Creep Rate for FC-5 Lime at Various Stages of Aging







79






7.00
6.00

.2 5.00
S 4.00 U Saturated
Unaged
3.00 -
OField Mix
u 2.00 Unaged
1.00 -
0.00
FC-5 G FC-5 L NOVACHIP




Fig. 7.17 Comparison of ER between unaged Saturated and Field mix





10.00
9.00 -
8.00 -
7.00 -
6.00 -
EC- Un aged
5.00 -
OAged
w 4.00 -
U-
3.00 -
2.00 -
1.00 -
0.00
FC-5 G FC-5 L NOVACHIP
(25% Red. In Air (25% Red. In Air (50% Red. In Air
Voids) Voids) Voids)


Fig. 7.18 FE for Saturated Mix











10.00
9.00
8.00
i 7.00


naiur Unaged
5.00 -
(u. 0 Aged

S 3.00 -
2.00 -
1.00 -
0.00
FC-5 G FC-5 L NOVACHIP
(25% Red. In Air (25% Red. In Air (50% Red. In Air
Voids) Voids) Voids)




Fig. 7.19 DCSE for Saturated Mix





Failure Strain(Saturated)

6000.0
5000.0 -
4000.0 -
4w 3000.0 Unaged
OAged
2000.0 -
1000.0 -
0.0
FC-5 G FC-5 L NOVACHIP
(25% Red. In Air (25% Red. In Air (50% Red. In Air
Voids) Voids) Voids)


Fig. 7.20 Failure Strain for Saturated Mix











7.00

6.00 -

o 5.00 -

z 4.00 -

| 3.00 -

W 2.00 -

1.00 -

0.00
FC-5 G FC-5 L NOVACHIP
(25% Red. In Air (25% Red. In Air (50% Red. In Air
Voids) Voids) Voids)


Fig 7.21 ER for Saturated Mix


2.50

L 2.00

. 1.50


1.00

0 0.50

0.0I-
0.00


FC-5 G FC-5 L NOVACHIP
(25% Red. In Air (25% Red. In Air (50% Red. In Air
Voids) Voids) Voids)


Fig. 7.22 Tensile Strength for Saturated Mix


SUnaged
O Aged


* Unaged
DAged


























FC-5 G FC-5 L
(25% Red. In Air (25% Red. In Air
Voids) Voids)


* Unaged
OAged


NOVACHIP
(50% Red. In Air
Voids)


Fig. 7.23 Resilient Modulus for Saturated Mix





5 8.00E-08
7.00E-08
U,
-L 6.00E-08
^ 5.00E-08
4.00E-08
i 3.00E-08 -
z 2.00E-08 -
1.00E-08 -
*5 O.OOE+00 -
FC-5 G FC-5 L NOVACHIP
(25% Red. In Air (25% Red. In Air (50% Red. In Air
Voids) Voids) Voids)


Fig. 7.24 Creep Strain Rate for Saturated Mix


* Unaged

O Aged









7.1.5 Moisture Conditioning

The FC-5 mixes were subjected to moisture conditioning as outlined in AASHTO-

T283. The procedure is follows:

For each of the mixes, the pills were compacted to 50 gyrations. The pills were then

placed in a vacuum bath to remove the entrapped air. Then, they were conditioned by

placing them in a water bath at 600C for 24 hours. After this, the samples were again

conditioned in a 25C water bath for two hours. Then water was allowed to drain down

for 36 hrs before cutting the samples. The samples were then tested and the results are as

presented below:

In case of FC-5 Granite, two of three samples seemed to have failed in the creep

test itself. The creep test results for the samples are presented in Fig 7.251. As evident

from the figures, for a load as low as 4 pounds, the samples were having a deformation

close to 900 micro inches. Infact at that low a load, there was no elastic response and the

creep curve did not seem to follow the power model. Hence, the creep rate for all the

samples was calculated using the straight-line portion of the creep deformation curve and

the results were compared to that of the unconditioned samples (Fig 7.25)

Clearly from Fig 7.26 it is evident that the creep rate for conditioned samples

increased significantly and that the specimens failed during creep test itself. Hence, the

strength test results are not truly representative of the behavior of the mix. Further, the

resilient modulus of the conditioned samples was lower as compared to unconditioned

samples (close to 3.6 M Pa). This clearly indicates that the samples had been damaged

during conditioning. Thus FC-5 Granite is highly sensitive to moisture conditioning.


1 Refer Appendix C for all the creep test results












0.0009
0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
0.0001
0


0 200 400 600
t (Sec)


0.0012

S0.001

c 0.0008

o 0.0006

E 0.0004

o 0.0002

0


0 200 400 600
Time (Sec)


800 1000 1200


800 1000 1200


Fig. 7.25 Creep Deformation Curve for Conditioned samples of FC-5 with Granite











3.500E-06

3.000E-06

2.500E-06

2.000E-06

1.500E-06

1.OOOE-06

5.000E-07

O.OOOE+00


G12 G21 G22 Gcon-l Gconl2 Gcon22
G1-2 2 cn1 cn2 Go2


Fig. 7.26 Creep Rate for Unconditioned and Conditioned samples FC-5 Granite

Moisture conditioning helps aging the mix and hence we see an increase in Mr and

tensile strength for conditioned FC-5 Limestone samples. Further, due to increased

absorption during conditioning, we see an increase in FE and ER for conditioned

samples. In addition, the creep rate decreases significantly for the conditioned samples.

The results are presented in Figures 7.27 through 7.32.


Unconditioned


Conditioned


Fig.7.27 FE after Conditioning for FC-5 with Limestone



















1.00
0.90
0.80 -
0.80

0.60 -
0.70



0.40 -
0.60
0.50
0.40
0.30
0.20
0.10
0.00


Unconditioned


Fig.7.28 DCSE after Conditioning for FC-5 with Limestone


3.00

2.50

2.00

1.50
-

1.00 -

0.50 -

0.00 -


Unconditioned


Fig.7.29 ER after Conditioning for FC-5 with Limestone


Conditioned


Conditioned


I







87







1200

1000

S800
U)
2 600
U
S400
.4-
200

0
Unconditioned Conditioned




Fig.7.30 Failure Strain after Conditioning for FC-5 with Limestone








10
9
8
7

5 5
4
3
2
1
0
Unconditioned Conditioned


Fig.7.31 Mr after Conditioning for FC-5 with Limestone