• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Fracture characteristics of asphalt concrete in Mode I, Mode II, and Mixed-mode

Dissertation
Author: Andrew Franz Braham
Abstract:
To this point, most research in asphalt concrete has focused on Mode I, or opening, cracking. However, a recent full-scale test section at the University of Illinois at Urbana-Champaign (UIUC) showed that cracks did not form in pure Mode I. If the cracks traveled vertically, the cracks could be considered a pure opening displacement, or Mode I, failure. However, the cracks did not form vertically, but instead traveled at an angle, indicating that there was also some sliding displacement, or Mode II, occurring. The combination of Mode I and Mode II can be referred to as Mixed-Mode. This research began with a review of existing Mode I literature for asphalt concrete. In order to develop a strong understanding of the Mode I behavior of asphalt concrete, 28 asphalt concrete mixtures were tested using the Disk-Shaped Compact Tension Test, which captures fracture energy (G f ). These mixtures included 11 different types of asphalt cement, three testing temperatures, two different types of aggregate, two different levels of air voids, and two different contents of asphalt cement. In addition, two Mode I fracture tests were compared: the Disk-Shaped Compact Tension Test and the Single-Edge Notch Beam Test. Three mixtures from the full-scale test sections discussed above were used to develop and validate a new Mode II test for asphalt concrete. No previous literature was found that attempted to capture Mode II fracture energy in asphalt concrete, so a test was developed from an extensive review of Mode II tests in the fields of metal, plastic, and PCC. Limited work has been performed in Mixed-Mode testing of asphalt concrete, but only Crack Mouth Opening Displacement (CMOD) was used. This research captured the Crack Tip Opening Displacement (CTOD) and Crack Tip Sliding Displacement (CTSD) in addition to CMOD. This allowed the comparison of three different values of fracture energy, or work, depending on the location and orientation of the data collection gages. Finally, the three asphalt concrete mixtures studied were simulated in Mode I to determine the total external work, with a break down of this work into the dissipated fracture energy, recoverable strain energy, and dissipated creep energy.

Table of Contents List of Figures vii List of Tables xii Chapter 1 -Introduction 01 1.1 Background 01 1.2 Fracture in Asphalt Concrete 05 1.3 Existing Mode I Fracture Testing in Asphalt Concrete 06 1.3.1 Single Edge Notch Beam [SE(B)] 06 1.3.2 Semi-Circular Bend [SC(B)] 14 1.3.3 SUPERPAVE Indirect Tension Test Fracture Parameter 18 1.3.4 Compact Tension Test 20 1.4 Objectives and Hypothesis 23 1.5 Research Scope 24 Chapter 2 - Fracture Characteristics of Asphalt Concrete in Mode 1 25 2.1 LTC Pooled Fund Study. 25 2.2 Disk-Shaped Compact Tension Testing [DC(T)] 26 2.3 Experimental Matrix 29 2.4 DC(T) Experimental Results 31 2.4.1 Asphalt Binder Type. 31 2.4.2 Aggregate Type 31 2.4.3 Temperature 34 2.4.4 Asphalt Content 36 2.4.5 Air Voids. 37 2.5 DC(T) Data Extrapolation 38 2.6 DC(T) Statistical Analysis 40 2.7 DC(T) versus SE(B) 41 2.8 CTOD versus CMOD 45 2.9 Mode I Summary and Conclusions 49 Chapter 3 - Fracture Characteristics of Asphalt Concrete in Mode II 51 3.1 Mode II Test Configurations 51 v

3.1.1 Double-Edge Notched 51 3.1.2 Shear Box 52 3.1.3 Arcan 53 3.1.4 Four Point Shear 54 3.2 Testing Configuration 55 3.3 MnROAD Test Results 59 3.4 ATLAS Test Results 63 3.5 Mode II Conclusions and Recommendations 72 Chapter 4 - Fracture Characteristics of Asphalt Concrete in Mixed-Mode 74 4.1 Mixed-Mode Test Configurations 74 4.1.1 Brazilian Disk with Angled Interior Notch 75 4.1.2 Compact Tension Shear 75 4.1.3 Single-Edge Notch Beam with Offset Notch 77 4.2 Mixed-Mode Testing of Asphalt Concrete 78 4.3 Specimen Fabrication and Data Collection 78 4.4 ATLAS Test Results 80 4.5 ATLAS Data Analysis and Discussion 85 4.6 Statistical Comparison for Mode I, Mode II, and Mixed-Mode 90 4.7 Energy Balance Simulations 91 4.8 Mixed-Mode Conclusions and Recommendations 98 Chapter 5 -Conclusions and Recommendations 100 5.1 Review of Objectives and Hypothesis 100 5.2 Recommendations 102 References 104 Appendix A - Data from Chapter 2 111 Appendix B - Data from Chapter 3 125 Appendix C - Data from Chapter 4 136 Curriculum Vitae 149 VI

List of Figures Figure 1.1 -ATLAS Test Sections 01 Figure 1.2 - Test Section's Pavement Structure 02 Figure 1.3 -Asphalt Concrete Cut 03 Figure 1.4 - Asphalt Concrete Reflective Cracking 04 Figure 1.5 - Mixed-Mode Cracking in Asphalt Concrete 05 Figure 1.6 - Single-Edge Notch Beam [SE(B)] Testing Configuration 06 Figure 1.7 - Single-Edge Notched Beam 12 Figure 1.8 - Semi-Circular Bend [SC(B)] Testing Configuration 14 Figure 1.9 - Semi-Circular Bend Test 15 Figure 1.10 - SUPERPAVE Indirect Tension Test Fracture Parameter Testing Configuration 18 Figure 1.11- Compact Tension Test (left) and Disc-Shaped Compact Tension [DC(T)] (right) Testing Configurations 20 Figure 1.12 - Disk-Shaped Compact Tension 21 Figure 2.1 -DC(T) Specimen (50mm thickness) .26 Figure 2.2 - Typical Fracture Energy Curves at High, Mid, and Low Testing Temperatures 29 Figure 2.3 - Effect of Aggregate Type on Fracture Energy. 32 Figure 2.4a - Typical Fracture Faces of Granite Mixtures, High Temperature on the Left, Low Temperature on the Right, (PG58-28, Unmodified, Source 2,4% Air Voids, 6.0% Asphalt Content) 33 Figure 2.4b - Typical Fracture Faces of Limestone Mixtures, High Temperature on the Left, Low Temperature on the Right, (PG58-40, SBS, 4% Air Voids, 6.9% Asphalt Content) 33 Figure 2.5 - Trends of Fracture Energy and Testing Temperature 34 Figure 2.6 - Fracture Energy Curves for Quasi-Brittle (PG64-22 with Granite Tested at -24°C) and Compliant (PG58-40 with +0.5%AC and Granite Tested at -18°C) Asphalt Concrete Mixtures 36 vn

Figure 2.7 - Effect of Asphalt Content on Fracture Energy 37 Figure 2.8 - Effect of Air Voids on Fracture Energy 38 Figure 2.9a - Full Load-CMOD Curve 39 Figure 2.9b - Exponential Extrapolation Technique 39 Figure 2.10 - Extraction of Field Cores versus Field Slabs 42 Figure 2.11 - SE(B) Specimen (75mm thickness) 43 Figure 2.12 - Low Testing Temperature CMOD Fracture Energy (J/m2) 43 Figure 2.13 - Intermediate Testing Temperature CMOD Fracture Energy (J/m2). 44 Figure 2.14 - High Testing Temperature CMOD Fracture Energy (J/m2) 44 Figure 2.15 - DC(T), Left, and SE(B), Right; Distance from CMOD Gage to Mechanically Inserted Notch 46 Figure 2.16 - Low Testing Temperature CTOD Fracture Energy (J/m2) 47 Figure 2.17 - Intermediate Testing Temperature CTOD Fracture Energy (J/m2) 48 Figure 2.18 - High Testing Temperature CTOD Fracture Energy (J/m2) 48 Figure 3.1 - Double-Edge Notch Test Configuration 52 Figure 3.2 - Shear Box Test Configuration 52 Figure 3.3 - Arcan Test Configuration (Sample on Left and Fixture on Right) 53 Figure 3.4 - Four Point Shear Test Configuration 54 Figure 3.5 - Initial Mode II Test Set-up 55 Figure 3.6 - Initial Mode II Schematic 56 Figure 3.7 - Load vs. CTSD Curve for Initial Mode II Test Set-up 56 Figure 3.8 - Work until Crack Initiation (WCTSD) 58 Figure 3.9 - Intermediate Mode II Test Set-up 59 Figure 3.10 - Load-Line Displacement Control Curves 60 Figure 3.11- Crack-Tip Sliding Displacement Control Curves 61 Figure 3.12 - Comparison of GCTSD for Crack-Tip Sliding Displacement Control Versus Load-Line Displacement Control at Three Temperatures 61 Figure 3.13 - Incorrect Gage Point Placement (Left), Correct Placement (Right, Zoomed) 63 viii

Figure 3.14-ATLAS Test Section Construction 63 Figure 3.15 - Final Mode II Test Set-Up 65 Figure 3.16 - Final Mode II Test Schematic 66 Figure 3.17 - Clip Gages for Mode II Test 67 Figure 3.18 - Mode II Peak Load of ATLAS Mixtures 68 Figure 3.19 - Mode II Work of ATLAS Mixtures 69 Figure 3.20 - Mode II Specimens After Testing: 9.5mm on Top, Transition in Middle, RCRI on Bottom 71 Figure 4.1 - Brazilian Disk with Angled Interior Notch 75 Figure 4.2 - Compact Tension Shear, Sample on the Left with Fixtures and Samples on the Right 76 Figure 4.3 - Fracture Parameters 77 Figure 4.4 - Beam Compactor with Mold Inserted (left) and Compaction Mold (right) 79 Figure 4.5 - Single Edge Notch Beam in Mode I Testing Configuration 79 Figure 4.6 - CTOD Gage and Gage Points on the Left; CTSD Gage and Gage Points on the Right (Not to Scale) 80 Figure 4.7 - Example Load vs. CMOD Curves Mixture 81 Figure 4.8 - Example Load vs. CTOD Curves Mixture 82 Figure 4.9 - Example Load vs. CTSD Curves Mixture 82 Figure 4.10 - Peak Load versus Off-Set Ratio 86 Figure 4.11 - WCMOD Work versus Off-Set Ratio 87 Figure 4.12 - WCTOD Work versus Off-Set Ratio 87 Figure 4.13 - WCTSD Work versus Off-Set Ratio 88 Figure 4.14 - WCTOD/WCMOD versus Off-Set Ratio 89 Figure 4.15 - WCTSD/WCMOD versus Off-Set Ratio 89 Figure 4.16 - WCTSD/WCTOD versus Off-Set Ratio 90 Figure 4.17 - Mesh for SE(B) Simulations in ABAQUS 92 Figure 4.18 - Energy Balance for 9.5mm Mixture 94 Figure 4.19 - Energy Balance for Transition Mixture 94 ix

Figure 4.20 - Energy Balance for RCRI Mixture 95 Figure 4.21 - Percent of External Work for 9.5mm Mixture 96 Figure 4.22 - Percent of External Work for Transition Mixture 96 Figure 4.23 - Percent of External Work for RCRI Mixture 97 FigureB.l -Trial 1, CTSD Control 126 Figure B.2 - Trial 2, CTSD Control 126 Figure B.3 - CTSD Control (0.7 mm/min), -6°C (Figure 3.11) 127 Figure B.4 - LLD Control (1.1 mm/min), -6°C (Figure 3.11) 127 Figure B.5 - CTSD Control (0.7 mm/min), -18°C (Figure 3.11) 128 Figure B.6 - LLD Control (1.1 mm/min), -18°C (Figure 3.11). 128 Figure B.7 - CTSD Control (0.7 mm/min), -30°C (Figure 3.11) 129 Figure B.8 - CTSD Control (0.7 mm/min), -30°C (Figure 3.11) 129 Figure B.9 - 9.5mm CTOD, -12°C (Table 3.1) 130 Figure B.10 - 9.5mm CTSD, -12°C (Table 3.1) 130 Figure B.l 1 - Transition CTOD, -12°C (Table 3.1) 131 Figure B.12 - Transition CTSD, -12°C (Table 3.1) 131 Figure B.13 - RCRI CTOD, -12°C (Table 3.1) 132 Figure B.14-RCRI CTSD, -12°C (Table 3.1) 132 Figure C.l - 9.5mm CMOD, 0mm Offset, -12°C (Table 4.1) 137 Figure C.2 - 9.5mm CTOD, 0mm Offset, -12°C (Table 4.1) 137 Figure C.3 - 9.5mm CTSD, 0mm Offset, -12°C (Table 4.1) 137 Figure C.4 - 9.5mm CMOD, 25mm Offset, -12°C (Table 4.1) 138 Figure C.5 - 9.5mm CTOD, 25mm Offset, -12°C (Table 4.1) 138 Figure C.6 - 9.5mm CTSD, 25mm Offset, -12°C (Table 4.1) 138 Figure C.7 - 9.5mm CMOD, 50mm Offset, -12°C (Table 4.1). 139 Figure C.8 - 9.5mm CTOD, 50mm Offset, -12°C (Table 4.1) 139 Figure C.9 - 9.5mm CTSD, 50mm Offset, -12°C (Table 4.1) 139 Figure CIO - 9.5mm CMOD, 75mm Offset, -12°C (Table 4.1) 140 Figure C.ll -9.5mm CTOD, 75mm Offset, -12°C (Table4.1) 140 x

Figure C.12 - 9.5mm CTSD, 75mm Offset, -12°C (Table 4.1) 140 Figure C.13 - Transition CMOD, 0mm Offset, -12°C (Table 4.2) 141 Figure C.14 - Transition CTOD, 0mm Offset, -12°C (Table 4.2) 141 Figure C.15 - Transition CTSD, 0mm Offset, -12°C (Table 4.2) 141 Figure C.16 - Transition CMOD, 25mm Offset, -12°C (Table 4.2) 142 Figure C.17 - Transition CTOD, 25mm Offset, -12°C (Table 4.2) 142 Figure C.18 - Transition CTSD, 25mm Offset, -12°C (Table 4.2) 142 Figure C.19 - Transition CMOD, 50mm Offset, -12°C (Table 4.2) 143 Figure C.20 - Transition CTOD, 50mm Offset, -12°C (Table 4.2) 143 Figure C.21 -Transition CTSD, 50mm Offset, -12°C (Table 4.2) 143 Figure C.22 - Transition CMOD, 75mm Offset, -12°C (Table 4.2) 144 Figure C.23 - Transition CTOD, 75mm Offset, -12°C (Table 4.2) 144 Figure C.24 - Transition CTSD, 75mm Offset, -12°C (Table 4.2) 144 Figure C.25 -RCRI CMOD, 0mm Offset, -12°C (Table 4.3) 145 Figure C.26 - RCRI CTOD, 0mm Offset, -12°C (Table 4.3) 145 Figure C.27 -RCRI CTSD, 0mm Offset, -12°C (Table 4.3). 145 Figure C.28 - RCRI CMOD, 25mm Offset, -12°C (Table 4.3) 146 Figure C.29 - RCRI CTOD, 25mm Offset, -12°C (Table 4.3) 146 Figure C.30 - RCRI CTSD, 25mm Offset, -12°C (Table 4.3) 146 Figure C.31 - RCRI CMOD, 50mm Offset, -12°C (Table 4.3) 147 Figure C.32 - RCRI CTOD, 50mm Offset, -12°C (Table 4.3) 147 Figure C.33 - RCRI CTSD, 50mm Offset, -12°C (Table 4.3) 147 Figure C.34 - RCRI CMOD, 75mm Offset, -12°C (Table 4.3) 148 Figure C.35 - RCRI CTOD, 75mm Offset, -12°C (Table 4.3) 148 Figure C.36 - RCRI CTSD, 75mm Offset, -12°C (Table 4.3) 148 xi

List of Tables Table 1.1 - Summary of the Single-Edge Notched Beam [SE(B)] Test 13 Table 1.2 - Summary of the Semi-Circular Bend [SC(B)] Test 17 Table 1.3 - Summary of Superpave IDT Fracture Parameter 19 Table 1.4 - Summary of Compact Tension Test 22 Table 2.1 - Aggregate Gradations, Percent Passing 30 Table 2.2 - Significance of Gf for Various Mixture Parameters 40 Table 2.3 - Comparison DC(T) and SE(B) 45 Table 2.4 - CTOD/CMOD Comparison 49 Table 3.1 - Work Summary for Mode II 69 Table 4.1 - 9.5mm Mixture Crack Initiation Work Summary 84 Table 4.2 - Transition Mixture Crack Initiation Work Summary 84 Table 4.3 - RCRI Mixture Crack Initiation Work Summary. 84 Table 4.4 - Gamma Values 85 Table 4.5 - Coefficient of Variation for Mode I and Mixed-Mode Tests 91 Table 4.6 - Coefficient of Variation for Mode II Tests 91 Table 4.7 - Mode I Fracture Energy Summary. 92 Table B.l - Creep and Tensile Strength for 9.5mm Mixture. 133 Table B.2 - Creep and Tensile Strength for RCRI Mixture 134 Table B.3 - Creep and Tensile Strength for Transitional Mixture 135 xn

Chapter 1 - Introduction 1.1 Background To this point, most research in asphalt concrete has focused on Mode I, or opening, cracking. This work is reviewed in Chapter 1.2. However, a recent full-scale test section at the University of Illinois at Urbana-Champaign (UIUC) showed that cracks did not form in pure Mode I. Figure 1.1 shows a picture of the full-scale test section. Figure 1.1 - ATLAS Test Sections The full scale test section was constructed in Fall 2007, at the Advanced Transportation Research Engineering Laboratory in Rantoul, IL. ATREL is a research, educational, and testing laboratory within the Department of Civil and Environmental Engineering at UIUC. The full scale test section was part of a project sponsored by the National Science Foundation (NSF) and private industry (Koch Materials Company and SemMaterials Company). There were four, 75 foot long test sections constructed for the 1

ATLAS testing program reported by Dave et al. (2007,2008a, 2008b). Each test section consisted of a Portland Cement Concrete (PCC) existing pavement layer with various asphalt concrete overlays. This configuration promoted reflective cracking in the asphalt concrete layers over the joints between the PCC slabs. Reflective cracking is the movement of cracks from these joints up and through the overlaying asphalt concrete pavement. All four test sections consisted of a subgrade with a material separation fabric placed on top. A 304mm granular subbase was placed with edge drains running transversely along the pavement. The 150mm thick PCC pavement was constructed with three joint load transfer conditions. Figure 1.2 shows a schematic of the pavement layers. A brief summary of the details of each test section follows. Asphalt Concrete 100mm i' Portland Cement Concrete 150mm Granular Subbase 304mm Subgrade Figure 1.2 - Test Section's Pavement Structure Section 1 was a control section, with both doweled joints and saw-cut joints. A standard 100mm thick Illinois overlay, 9.5mm Nominal Maximum Aggregate Size (NMAS) asphalt concrete with a PG64-22 binder, was placed on top of the PCC pavement layer. Placing this standard overlay over a PCC surface represents a typical rehabilitation strategy in Illinois. Section 2 was termed the "interlayer" section, as the asphalt concrete overlay consisted of a 25mm Reflective Crack Relief Interlayer (RCRI), followed by (e.g., topped with) the 75mm standard overlay. The RCRI was a 4.75mm NMAS asphalt concrete with a highly polymer modified, proprietary binder. The RCRI is intended to reduce reflective cracking because of its highly ductile and compliant nature. 2

Section 3 explored the benefits of using a second, 'transitional' interlayer in conjunction with the traditional interlayer just described. Here, the asphalt concrete overlay consisted of a 25mm RCRI, with a 25mm "transitional" interlayer (this is the mixture termed "transition" in Chapters 3 and 4). The transition interlayer was a 4.75mm NMAS mixture with a blend of the polymer modified binder (66%) and PG64-22 (33%). A 50mm standard overlay was placed on top of these two layers. Section 4 has not been tested yet, so it will not be discussed further in this chapter. As mentioned earlier, if pure Mode I cracking were to occur, reflective cracks would propagate vertically from the PCC joints up through the asphalt layer. However, after testing three of the pavement sections, it was observed that cracks were not propagating in a pure Mode I fashion. The ATLAS test section was closely monitored to observe crack patters in the asphalt concrete layers. In Figure 1.1, a slice of the asphalt concrete layers was cut so that crack patterns could be observed. Figure 1.3 is a zoomed image of one of the cuts. Figure 1.3 - Asphalt Concrete Cut 3

For the accelerated testing, the loading level varied from 2- 30 kips (~9 - 133 kN), with approximately 20,000 passes total. Zooming in on Figure 1.3, as seen in Figure 1.4, provides a view of the actually cracking after the accelerated loading had been applied. Figure 1.4 - Asphalt Concrete Reflective Cracking There are two interesting observations from the crack pattern in Figure 1.4. First, the cracks initiate above the 25mm RCRI layer that is directly above the PCC surface. This indicates that the cracking "jumped" the highly ductile and compliant RCRI layer, and formed in the more brittle asphalt concrete layers. The second observation is that the cracks do not travel vertically from the PCC joint. If they did travel vertically, only Mode I, or opening, would be occurring. However, traveling at an angle indicates that there is also some sliding displacement, or Mode II, occurring. The combination of Mode I and Mode II in a material can be referred to as Mixed-Mode. Neither Mode II nor Mixed-Mode behavior has been investigated in any detail in asphalt concrete. This research begins with a review of existing Mode I literature for asphalt concrete, further investigates Mode I behavior of asphalt concrete, and begins the investigation of Mode II and Mixed-Mode behavior. In addition, since three very different mixtures were investigated in the ATLAS test section, these three mixtures are 4

simulated in a Single-Edge Notch Beam configuration to break down the fracture energy, or external work, into dissipated fracture energy, recoverable strain energy, and dissipated creep energy. 1.2 Fracture in Asphalt Concrete There are three types of fracture: Mode I, Mode II, and Mode III (Anderson, 2005). These types of fracture can be translated into types of cracking. Mode I cracking is an opening cracking; Mode II is an in-plane shear cracking; and Mode III is an out-of-plane shear cracking. One mechanism of Mode I cracking in asphalt concrete results from thermal cracking. As asphalt concrete cools, it contracts. When these thermal stresses exceed the strength of the pavement, it cracks. However, a moving wheel load (traffic on the pavement) introduces Mode II loading. This mode is especially prevalent in composite pavements, where asphalt concrete is laid over an existing Portland Cement Concrete (PCC) pavement. As the wheel moves away from the existing PCC joint, it generates a shearing action in the asphalt concrete overlay. This is called reflective cracking, and is a combination of bending of the lower layers, the wheel load, and thermal effects on the pavement, as seen in Figure 1.5. Asphalt Concrete PCC Slab Base Course Subgrade Figure 1.5 - Mixed-Mode Cracking in Asphalt Concrete 5

The combination of the influence of bending of lower layers, thermal effects, and a wheel load cracking is called Mixed-Mode cracking. This Mixed-Mode cracking has been studied extensively in metals, rocks, and PCC (see Chapter 4), but the majority of the work in asphalt concrete has been only in Mode I fracture testing. The next section discusses the development of Mode I fracture testing in asphalt concrete. 1.3 Existing Mode I Fracture Testing in Asphalt Concrete Most of the literature dealing with asphalt concrete fracture has dealt with Mode I fracture. There are four main testing configurations developed to date are: the Single- Edge Notch Beam [SE(B)], the Semi-Circular Bend [SC(B)], the SUPERPAVE Indirect Tension Test fracture parameter, and the Disk-Shaped Compact Tension Test [DC(T)]. The development of these tests will be discussed in the following four sections. Each section has a summary table at the end for quick reference. 1.3.1 Single Edge Notch Beam [SE(B)] The first section will discuss research in asphalt concrete using the Single Edge Notch Beam [SE(B)] configuration. Figure 1.6 shows the general, 3-point bend testing setup for all SE(B) tests. Figure 1.6 - Single-Edge Notch Beam [SE(B)] Testing Configuration 6

The first research on the application of fracture mechanics concepts to asphalt mixtures (Moavenzadeh, 1967) was performed on a SE(B). Moavenzadeh explored the strain- energy release rate, which is shown in Equation 1.1 ° = E xS"xhx f^d^ E

dc aC =AxKn i-Kj , Equation 1.2 where dc/dN = rate of crack propagation A and n = material constants K = stress intensity factor. The stress field, Kc, in the vicinity of the crack tip is show in Equation 1.3 K*=a2nx(l-ju2)xhxf(cf/d) .Equation 1.3 where on - MJb X (d - cf M = bending moment = Pxl/4 P = load at fracture b = width of beam d = depth of beam c = notch length l = span \x = Poisson's ratio h = d- c f(cf/d) = function of the crack depth at failure (cf) to depth (d) ratio. Their analysis concluded that as the testing temperature increased (from -5°C to 25 °C), the influence of creep increased, which decreased the n value for a beam. However, if the creep effect was ignored, the constant n approximately equaled four, which was the value predicted by Paris' analysis. Many more studies over the following years investigated the fracture toughness of asphalt concrete using Linear Elastic Fracture Mechanics (LEFM). Ramsamooj et al. (1991) 8

tested 46 different mixtures at temperatures ranging from 16-35°C using Paris' Law. In order to simulate the underside of an asphalt concrete slab with vehicle loading at 50 km/hr they applied a stress rate of 6890 kPa/s. One of their major conclusions was the fatigue life of their three asphalt concrete mixtures could be predicted from simple fracture tests. Kim et al. (1997) investigated the fracture toughness of asphalt concrete from -5 to -30°C in 5°C steps. Fracture toughness was calculated using Equation 1.4 KIC = 1.14X an X Ja~X f(ac) Equation 1.4 where Kc = fracture toughness on=6xM/BxW M=(P + wS/2)x(S/4) B = thickness W = height P = maximum applied load w = weight per unit length of beam S = span ac = effective crack length f(ae) = function of effective crack length to beam height ratio. They used penetration grade 85-100 asphalt cement with granite and limestone aggregate. They found that toughness increased from -5 to -15°C and then decreased down to -30°C. They also indicated that granite showed a slightly better resistance to fracture than limestone. Mobasher et al. (1997) compared stress intensity factors and toughness, at -1 and -7°C, of conventional asphalt cement to asphalt cement with rubber (asphalt rubber). They found that they asphalt rubber mix has a lower modulus but a higher toughness than conventional asphalt cement. Hossain et al. (1999) also investigated the effect of rubber content on fracture energy. Fracture energy was calculated through Equation 1.5 9

r _W0+2xPwxS °> ~ /, \ Equation 1.5 (b-a)xt where GF = fracture energy Wo = area under load/displacement curve, not including self weight of beam Pw = equivalent load induced by self-weight of beam 8 - total deflection b = specimen height a = initial notch length t = thickness of specimen. Using three rubber contents (19,22,24%), three asphalt cement contents (6, 7.5,9%), and two temperatures (5,25°C), they found higher values of fracture energy from higher binder contents, irrespective of rubber content. Marasteanu et al. (2002) used ASTM E399, Linear-Elastic Plane-Strain Fracture Toughness Kic of Metallic Materials, to find the fracture toughness of asphalt concrete. Equation 1.6 was used for the three-point bend test K-i ~ ~ j^372~X f\a ) Equation 1.6 where Ki = fracture toughness F = load applied S = specimen span B = specimen thickness W = specimen width f(a) = function of specimen crack size and width. 10

All tests were run with air voids of approximately 10 percent and loading records indicated that nonlinear behavior was more pronounced at -18°C than at -34°C, which means that this asphalt mixture would be more resistant to fracture at -18°C than at - 34°C. Using the compliance method, they obtained the fracture resistance curve, called the R-curve, of the asphalt mixture. Initial results showed that as the crack length increased, the fracture toughness increased as well. Here, it is important to remember that the fracture toughness has little application to asphalt concrete (unless tested at very low temperatures or in large structures) as asphalt concrete displays non-linear, viscoelastic behavior. A few studies characterized fracture using the strain energy release rate based on Elastic- Plastic Fracture Mechanics (EPFM). This is found by calculating the J integral, Jic, as demonstrated in Equation 1.7 JIc=2x UTlb X (d - a) Equation 1.7 where Jic = critical strain energy release rate Uj/b = total strain energy (area under load/displacement plot) d = specimen height a = notch length Dongre et al. (1989) compared 12 asphalt cements with the same aggregate at four temperatures (-21 to +16°C). At low temperatures, Jic, the critical J integral, was sensitive to asphalt concrete stiffness, asphalt cement grade, and asphalt cement source, while Kic (found through Equation 1.6) was not. Bhurke et al. (1997) compared multiple polymer additives to binder using the J integral. Two modifiers were tested at three polymer loading levels. They found a parabolic relationship, where an increase in polymer initially increased strain energy, but after a point it decreased the strain energy. This indicated an optimal polymer loading to maximize strain energy. 11

Figure 1.7 - Single-Edge Notched Beam More recently, Wagoner et al. (2005a, 2005b), determined that the Single Edge Notched Beam, SE(B), was the most promising fracture test based on test control method as seen in Figure 1.7. They came to this conclusion because of the SE(B)'s crack front development, test repeatability, test temperature, and ability to capture mixed mode fracture. However, they did not find a standard test method or analysis for asphalt concrete. They compared three different nominal maximum aggregate sizes with three binders and found that the polymer modified mixture with the smallest aggregate size gave the highest fracture energy values. They successfully measured mixed-mode fracture by offsetting the notch on the bottom of the specimen, which is discussed further in Chapter 4. Finally, Artamendi et al. (2007) compared the effect of specimen geometry and loading rate using fracture toughness (Equation 1.4). They compared the SE(B) with the Semi- Circular Bend SC(B) configuration at 1, 5, and 10 mm/min with Stone Mastic Asphalt and a Dense Bitumen Macadam mixtures. They investigated maximum load, the critical load, the stress intensity factor and the fracture energy. For the two mixtures they tested, the SE(B) gave lower fracture toughness values than the SC(B), and fracture toughness increased with an increasing loading rate. 12

Full document contains 165 pages
Abstract: To this point, most research in asphalt concrete has focused on Mode I, or opening, cracking. However, a recent full-scale test section at the University of Illinois at Urbana-Champaign (UIUC) showed that cracks did not form in pure Mode I. If the cracks traveled vertically, the cracks could be considered a pure opening displacement, or Mode I, failure. However, the cracks did not form vertically, but instead traveled at an angle, indicating that there was also some sliding displacement, or Mode II, occurring. The combination of Mode I and Mode II can be referred to as Mixed-Mode. This research began with a review of existing Mode I literature for asphalt concrete. In order to develop a strong understanding of the Mode I behavior of asphalt concrete, 28 asphalt concrete mixtures were tested using the Disk-Shaped Compact Tension Test, which captures fracture energy (G f ). These mixtures included 11 different types of asphalt cement, three testing temperatures, two different types of aggregate, two different levels of air voids, and two different contents of asphalt cement. In addition, two Mode I fracture tests were compared: the Disk-Shaped Compact Tension Test and the Single-Edge Notch Beam Test. Three mixtures from the full-scale test sections discussed above were used to develop and validate a new Mode II test for asphalt concrete. No previous literature was found that attempted to capture Mode II fracture energy in asphalt concrete, so a test was developed from an extensive review of Mode II tests in the fields of metal, plastic, and PCC. Limited work has been performed in Mixed-Mode testing of asphalt concrete, but only Crack Mouth Opening Displacement (CMOD) was used. This research captured the Crack Tip Opening Displacement (CTOD) and Crack Tip Sliding Displacement (CTSD) in addition to CMOD. This allowed the comparison of three different values of fracture energy, or work, depending on the location and orientation of the data collection gages. Finally, the three asphalt concrete mixtures studied were simulated in Mode I to determine the total external work, with a break down of this work into the dissipated fracture energy, recoverable strain energy, and dissipated creep energy.