ABSTRACT
XUE, LEI. Investigation on the Debonding Potential of Geosynthetic-Reinforced Asphalt Pavements. (Under the direction of Dr. Y. Richard Kim).
Asphalt overlays are widely used by highway engineers as a quick and reliable rehabilitation technique to treat distressed pavements. These freshly prepared overlays often crack within no-time under the repeated vehicular loading and temperature fluctuations caused by day to night changes and seasonal variations. The phenomenon of crack propagation through the new overlay from the existing cracks in the underlying pavement structure is known as reflective cracking. Many interlayer reinforcement techniques are employed to mitigate reflective cracking, but geosynthetics products are gaining attention nowadays due to their ease in installation, low cost, and wide availability. However, proper bonding between adjacent asphalt layers is a significant concern that allows the pavement structure to act monolithically in resisting vehicular and thermal loads. A weak bond between the layers and the geosynthetic product which meets the necessary tensile requirements for paving could eventually end in premature failure following a reduction in the service life of asphalt pavements. Therefore, the proper selection criterion for geosynthetic products to meet the different pavement conditions need utmost emphasis. The evaluation of interface bond strength is considered a quick and easy way to measure the debonding resistance. In this study, the monotonic shear test was performed using Modified Asphalt Shear Tester (MAST) under different confining pressures and temperatures on geosynthetic reinforced and unreinforced specimens. Thereby, with the aid of the time-temperature superposition principle, the shear bond strength prediction models were developed. These models rank the bonding and cracking potentials of geosynthetic products in pavement structure by predicting the shear strength at the layer interface and comparing against the typical pavement stress conditions mimicked using a three-dimensional finite element software package for moving load analysis, FlexPAVE™ version 1.1. According to MAST results, the presence of any type of geosynthetic products at any testing conditions reduces the interface shear strength (ISS) and increases the chances of debonding. The geosynthetic-reinforced specimens are categorically classified into two based on the shear strength mastercurves. Paving composite#1 (PC#1) and paving grid (PaG) display the
higher shear strength among the geosynthetic-reinforced specimens, while paving mat (PM), paving fabric (PF), and paving composite#2 (PC#2) show lower shear strength in comparison to the former category. The effect of the tack coat application rates on the ISS was determined using the statistical tool, Analysis of Covariance (ANCOVA). The statistical analysis suggests that the tack coat application rate of PM and PF have a significant effect on ISS. Further to which the Tukey’s honest significance or Tukey’s HSD analysis shows significant effect on ISS of PM and PF when the tack coat application rate changes from dry to optimum. Also, a similar significant effect on PF’s ISS is found when the application rate changes from optimum to wet, but not for PM. For the same type of geosynthetic-reinforced specimen, the shear strength decreases with an increase in temperature. Note that the effect of geosynthetic types on the ISS is evident at the low temperature (23C) but is nullified at the high temperature (54C). The increase in confining pressure from 172.37 kPa (25 psi) to 482.63 kPa (70 psi) results in an increase in ISS of PC#1 and PM. Maximum shear ratio analysis using FlexPAVE™ shows that the highest debonding potential among the simulation conditions in this study occurs in thick overlay structure, high temperature, low speed, thin overlay thickness. Consequently, the selection criterion for geosynthetic paving products that are safe against debonding is proposed. The recommended ISS testing condition is at 50C, 5.08 mm/min (0.2 in./min) actuator displacement rate, and 275.8 kPa (40 psi) confining pressure. The minimum required shear strength for the geosynthetic- reinforced specimen at this condition is 305 kPa (44 psi) for acceptance.
© Copyright 2020 by Lei Xue
All Rights Reserved
Investigation on the Debonding Potential of Geosynthetic-Reinforced Asphalt Pavements
by Lei Xue
A thesis submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the degree of Master of Science
Civil Engineering
Raleigh, North Carolina 2020
APPROVED BY:
______Dr. Y. Richard Kim Dr. Cassie Castorena Committee Chair
______Dr. Shane Underwood
ii DEDICATION
To Zhang Chunjie and Liu Zhaoquan, amor vincit omnia.
iii BIOGRAPHY
Lei Xue was born and grew up in Tibetan Plateau, Qinghai, China. He enjoyed his childhood and teenage life with his family. After graduation from Xining No.2 senior high school, he moved to Beijing. He received his bachelor’s degree in Transportation Engineering and Law from Beijing University of Technology. During his senior year, he decided to apply to North Carolina State University for the Master’s in Civil Engineering under the advice of Dr. Richard Kim.
iv ACKNOWLEDGMENTS
I would like to express my gratitude to my advisor Dr. Richard Kim, for his valuable guidance, support, and encouragement. A sincere thank you to my committee members Dr. Cassie Castorena and Dr. Shane Underwood. I would like to express my thank to Dr. Castorena for her support during my 2020 fall semester. I would also like to thank lab manager Ernie Song for cultivating a good laboratory environment. I would like to extend my thanks to the research group members and my friends. I would like to express my thank to Dr. Nithin Sudarsanan. It is my pleasure to work with you. I always impressed by your passion and enthusiasm for research. This dissertation finished during the COVID-19 globe pandemic. I would like to thank Chi Tian for her patience and love. I would like to thank my family, Liu Yu, Xue Yanxiang, and Niu Shuqing, for your unweaving love and encouragement.
v TABLE OF CONTENTS
LIST OF TABLES ...... vii LIST OF FIGURES ...... viii 1. INTRODUCTION ...... 1 1.1 Background ...... 1 1.2 Research Objectives ...... 1 1.3 Thesis Organization...... 2 2. LITERATURE REVIEW ...... 3 2.1 Reflective Cracking ...... 3 2.2 Function of Geosynthetic ...... 4 2.2.1 Reinforcing ...... 4 2.2.2 Stress Relieving ...... 4 2.2.3 Water Barrier ...... 4 2.3 Debonding Problem...... 5 2.4 Factors Influencing the Bonding ...... 5 2.4.1 Tack Coat Type ...... 5 2.4.2 Tack Coat Application Rate ...... 6 2.4.3 Curing Time ...... 7 2.4.4 Surface Texture ...... 7 2.5 Test Methods ...... 7 2.5.1 Ancona Shear Testing Research and Analysis (ASTRA) Device ...... 9 2.5.2 Louisiana Interlayer Shear Strength Tester (LISST) ...... 10 2.5.3 Sapienza Direct Shear Testing Machine (SDSTM) ...... 10 2.5.4 Advanced Shear Tester (AST) ...... 11 2.5.5 Modified Asphalt Shear Tester (MAST) ...... 12 2.6 Bonding of Geosynthetic-Reinforced Interlayer ...... 14 2.7 Critical Summary ...... 16 3. EXPERIMENTAL PLAN ...... 17 3.1 Materials ...... 17 3.1.1 Asphalt Mixture ...... 17 3.1.2 Tack Coat ...... 23 3.2 Specimen Fabrication ...... 27 3.2.1 Compaction ...... 27
vi 3.2.2 Geosynthetic Material Preparation ...... 28 3.2.3 Interface Installation ...... 31 3.2.4 Tack Coat Application ...... 35 3.2.5 Sample to Specimen Fabrication ...... 41 3.2.6 Air Void Study on MAST Specimen ...... 44 3.3 Testing Methodology ...... 45 3.3.1 Modified Asphalt Shear Tester ...... 45 3.3.2 Digital Image Correlation ...... 50 4. PAVEMENT RESPONSE ANALYSIS ...... 54 4.1 Parameters Used in FlexPAVE™ Simulation ...... 54 4.1.1 Structure Information ...... 54 4.1.2 Material Parameters for Each Pavement Layers ...... 56 4.1.3 Climate Data ...... 57 4.1.4 Traffic Data ...... 57 4.1.5 Tire-Pavement Contact Pressure Configuration ...... 57 4.1.6 Shear Traction Due to Tire Braking Condition...... 58 4.2 FlexPAVE™ Analysis Output ...... 58 4.2.1 Stress and Strain Distribution ...... 58 5. INTERFACE SHEAR STRENGTH ...... 62 5.1 Shear Strain Rate ...... 62 5.2 Interface Shear Strength ...... 63 5.2.1 Effects of Geosynthetic Interlayer Type on Interfacial Shear Strength ...... 63 5.2.2 Effects of Tack Coat Application Rate on Interface Shear Strength ...... 65 5.2.3 Effects of Temperature on Interface Shear Strength ...... 70 5.2.4 Effects of Confining Pressure on Interface Shear Strength ...... 70 5.3 Statistical Analysis on the Effect Tack Coat Application Rate ...... 73 5.4 Shear Strength Prediction Model ...... 75 5.5 Shear Ratio and Bonding Failure Definition ...... 76 5.6 Acceptance Criterion for Debonding Resistant Geosynthetic Products...... 89 6. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE STUDY ...... 91 REFERENCES ...... 94
vii LIST OF TABLES
Table 2.1 NCHRP report 712 tack coat application rate (Mohammad et al. 2012)...... 6 Table 2.2 FHWA tech brief tack coat application rate (Federal Highway Administration 2016)...... 6 Table 3.1 Linear viscoelastic properties of the test mixture...... 20 Table 3.2 Summary of tack coat application rates for geosynthetic products used in this study...... 24 Table 3.3 Averaged time-temperature shift factor function coefficients for PG 64-22...... 25 Table 3.4 Nomenclature details of different types of geosynthetics...... 29 Table 3.5 Properties of geosynthetic products...... 30 Table 3.6 MAST specimens: air void content...... 44 Table 3.7 Verification of air void study results...... 44 Table 4.1 Prony coefficients for relaxation modulus...... 56 Table 5.1 ANCOVA analysis summary...... 73 Table 5.2 Tukey HSD analysis for PM...... 74 Table 5.3 Tukey HSD analysis for PF...... 74 Table 5.4 Coefficients for PC#1 and PM shear strength prediction models...... 76 Table 5.5 MSR location information summary...... 90
viii LIST OF FIGURES
Figure 2.1 Mechanism of reflective cracking (Nithin et al. 2015)...... 3 Figure 2.2 Stress at interlayer caused by moving traffic (Raab and Partl 2004)...... 5 Figure 2.3 Fracture mechanics crack mode...... 8 Figure 2.4 Shear stress distribution (a) Direct Shear Test; (b) Simple Shear Test (Raab et al. 2009)...... 9 Figure 2.5 Ancona Shear Testing Research and Analysis (ASTRA) Device (Pasquini et al. 2015)...... 10 Figure 2.6 Louisiana interlayer shear strength tester (LISST) (Mohammad et al. 2018)...... 10 Figure 2.7 Sapienza Direct Shear Testing Machine (SDSTM) (Tozzo et al. 2014)...... 11 Figure 2.8 Advanced Shear Tester (AST) (Zofka et al. 2015)...... 11 Figure 2.9 Modified asphalt shear tester (MAST) (Cho et al. 2017b)...... 13 Figure 2.10 Typical MAST test result...... 13 Figure 2.11 Shear ratio concept...... 14 Figure 3.1 Aggregate gradation of RAP-40 mixture...... 17 Figure 3.2 (a) Removing AC loose mix with cloth bag from the bucket (b) Removing the AC lump from the cloth bag and (c) the loose mix lump inside the metal bucket...... 18 Figure 3.3 Separation pans and cloth bags for making well mix AC...... 19 Figure 3.4 The linear relation between the air voids and weight of gyratory samples...... 20 Figure 3.5 Dynamic modulus values for RAP-40...... 21 Figure 3.6 Time-temperature shift factor function for RAP-40 mixture...... 22 Figure 3.7 Dynamic modulus mastercurve for RAP-40...... 23 Figure 3.8 (a) Dynamic shear modulus mastercurve and (b) shift factor response of PG 64-22...... 26 Figure 3.9 (a) Superpave gyratory compactor (Pine Test Equipment, Inc.) and (b) compaction molds ...... 27 Figure 3.10 Compaction procedure for double-layered MAST sample: (a) bottom layer fabrication, (b) bottom layer placement in hot mold with tack coat, and (c) completed MAST sample...... 28 Figure 3.11 Geosynthetic samples: (a) PC#1, (b) PC#2, (c) PaG, (d) PM, and (e) PF...... 29 Figure 3.12 Placement of geosynthetics in the field...... 31 Figure 3.13 Geosynthetic interlayer sample cutting process...... 32 Figure 3.14 (a) Tracing the cutting pattern, (b) completed template pattern, and (c) cutting/extracting the geosynthetic sample using a cloth cutter...... 32
ix Figure 3.15 Placement of Geosynthetic Interlayers (view towards bottom layer): (a) PC#1, (b) PC#2, (c) PaG, (d) PM, and (e) PF...... 33 Figure 3.16 Setting pressure application on PC#1 using metal rod: (a) bottom layer after tack coat application, (b) liquified asphalt binder after placing the bottom layer with tack coat in the oven at 145C for two (2) minutes, and (c) the setting pressure application over the PC#1 by rolling the metallic rod...... 34 Figure 3.17 Tracking the geosynthetic-reinforcement placement direction (a) bottom layer of the sample (b) final cored specimen...... 35 Figure 3.18 Tack coat application process: (a) pouring hot binder from the metal canister with perforated lid and (b) spreading binder uniformly using a heat gun and the metal spatula...... 36 Figure 3.19 Non-uniformity found on tack coat applied bottom layer surface of MAST samples ...... 36 Figure 3.20 Details of: (a) test setup (b) the control panel (c) the hot spray gun...... 37 Figure 3.21 Hot spray gun: (a) components, (b) air spray nozzle, and (c) liquid nozzle...... 38 Figure 3.22 (a) Attaching liquid nozzle to hot cartridge, (b) pouring liquid asphalt into cartridge, (c) loading cartridge into heating chamber of gun, (d) cartridge with asphalt inside spray gun, (e) closing spray gun mouth, and (d) measuring the temperature at nozzle tip...... 39 Figure 3.23 (a) Sheet cover above gyratory sample, (b) small gap between the cover sheet and sample, and (c) application of hot asphalt using a hot spray gun...... 40 Figure 3.24 Tack coat applied to the bottom layer surface of MAST samples: (a) non-uniformity (metal canister) and (b) uniformity (hot spray gun)...... 41 Figure 3.25 Procedure for coring and cutting cylindrical specimens: (a) cored MAST sample, (b) trimming the top/bottom layer, and (c) finished specimens...... 42 Figure 3.26 Process for PVC pipe protection of geosynthetic-reinforced sample...... 43 Figure 3.27 Side view of geosynthetic-reinforced specimens with: (a) PC#1, (b) PC#2, (c) PaG, (d) PM, and (e) PF...... 43 Figure 3.28 Target and achieved air void content relationship for different layers...... 45 Figure 3.29 Illustrations of the Modified Advanced Shear Tester (MAST)...... 46 Figure 3.30 Gluing procedure for MAST specimen: (a) tightened bottom shoes on the gluing jig (b) application of glue on the bottom shoes, (c) specimen placement on the bottom shoes, (d) the upper shoe installation above the specimen, (e) specimen with all shoe in place, and (f) trimming the extra glue from the shoe edges...... 47 Figure 3.31 (a) Loading the MAST shoes with the specimen to the MAST jig, (b) installing the confining pressure plate with load cell,
x (c) placing the MAST over the MTS 810, (d) environmental chamber, (e) DIC test set-up, and (f) view through DIC camera...... 49 Figure 3.32 Loading configuration of MAST test set-up...... 50 Figure 3.33 DIC settings on the MAST to monitor on-specimen displacement...... 50 Figure 3.34 Digital image correlation process (Safavizadeh and Kim 2017)...... 51 Figure 3.35 Preparation of speckled paper: (a) spray painting the paper, (b) finished speckled paper, and (c) speckled paper on MAST shoe to track on-specimen displacement...... 51 Figure 3.36 Comparison of MTS actuator and DIC system displacements...... 52 Figure 4.1 Cross-section of thick pavement structure (X-axis: transverse direction and Y-axis: longitudinal/traffic direction) ...... 55 Figure 4.2 Pavement structure input: (a) thick structure, (b) intermediate structure, and (c) thin structure...... 55 Figure 4.3 FlexPAVE™ dual tires-pavement contact configuration...... 58 Figure 4.4 Stress distribution at the interface 1.5-inch deep: (a) normal stress and (b) shear stress...... 59 Figure 4.5 (a) Normal stress and (b) resultant shear stress distribution along the critical sections...... 60
Figure 4.6 Shear strain (a) yz and (b) zx...... 61 Figure 5.1 DIC strain rate calculation method (Chehab et al. 2002)...... 63 Figure 5.2 Mastercurve for different geosynthetic-reinforced asphalt specimens...... 64 Figure 5.3 Tack coat application rate effect on geosynthetic-reinforced specimen: (a) PC#1, (b) PaG, (c) PC#2, (d) PF, and (e) PM...... 66 Figure 5.4 Confining pressure effect on geosynthetic-reinforced specimens: (a) PC#1 and (b) PM...... 71 Figure 5.5 Correlation between shear strength and confining pressure: (a) PC#1 and (b) PM (The numbers in the legend indicate the reduced strain rate)...... 72 Figure 5.6 Pictorial representation of the effect of tack coat application rate on ISS: (a) PM, (b) PF...... 75 Figure 5.7 PC#1 prediction model validation at 172 kPa confining pressure...... 76
Figure 5.8 Shear strain output: (a) Eyz and (b) Exz ...... 78 Figure 5.9 PM Shear ratio, shear strength, and shear and normal stress levels in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN (9 kips), 1.6 km/hour (1 mph), 50C, at 3.81 cm (1.5 in.) depth, and braking condition...... 79
xi Figure 5.10 Shear ratio distribution under dual tires in 3D space (a) PC#1 condition, (b) PM condition and in 2D space contour (c) in PC#1 condition, and (d) PM condition...... 80 Figure 5.11 Effects of pavement structure on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: 40 kN, 1.6 km/hour, 50C, at 3.81 cm depth, and braking condition (a) PC#1 and (b) PM...... 82 Figure 5.12 Effects of pavement structure on MSR: 40 kN, 1.6 km/hour, 50C, at 3.81 cm depth, and braking condition...... 83 Figure 5.13 Effects of temperature on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN, 1.6 km/hour, at 3.81 cm depth, and braking condition (a) PC#1 and (b) PM...... 84 Figure 5.14 Effects of temperature on MSR: thick pavement, 40 kN, 1.6 km/hour, at 3.81 cm depth, and braking condition...... 85 Figure 5.15 Effects of overlay thickness on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN, 1.6 km/hour, 50C, and braking condition (a) PC#1 and (b) PM...... 86 Figure 5.16 Effects of overlay thickness on MSR: thick pavement, 40 kN, 1.6 km/hour, 50C, and braking condition...... 87 Figure 5.17 Effects of vehicle speed on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN, 50C, at 3.81 cm depth, and braking condition (a) PC#1; (b) PM...... 88 Figure 5.18 Effects of vehicle speed on MSR: thick pavement, 40 kN, 50C, at 3.81 cm depth, and braking condition...... 89 Figure 5.19 Critical condition schematic...... 89 Figure 5.20 Temperature relationship with DIC reduced strain rate at 5.08 mm/min MTS deformation rate...... 90
1 1. INTRODUCTION
1.1 Background One common form of rehabilitations used for pavement distress treatment is asphalt concrete overlays. Asphalt overlays are regarded as a quick and reliable rehabilitation technique to treat distressed pavement surfaces. The phenomenon of crack propagation through the new overlay from the underlying pavement structure is known as reflective cracking. Many interlayer reinforcement techniques are employed to mitigate reflective cracking, but geosynthetic products are gaining attention nowadays due to their ease in installation, low cost, and wide availability. The primary functions of the geosynthetic are reinforcing, stress-relieving, and waterproofing. The reinforcing function requires the geosynthetic material to have a significantly higher modulus than the surrounded asphalt. It can redirect crack propagation at the interlayer, which indefinitely delay or mitigate reflective cracking. Stress-relieving geosynthetic has lower stiffness and is capable of storing strain at low stress level. While the crack penetrates through the overlay, the geosynthetic acts as a barrier to prevent water infiltration and protects the underlying structure. The fully impregnated geosynthetic significantly reduces the water permeability. The proper installation, controlling overlay thickness, and overseeing compaction quality are required to achieve the functions of geosynthetic. However, proper bonding between adjacent asphalt layers is a major concern that allows the pavement structure to act monolithically in resisting vehicular and thermal loads. A weak bond between the layers and the geosynthetic product eventually ends in premature failure following a reduction in the service life of asphalt pavement. Therefore, the proper selection criterion for geosynthetic products to meet the different pavement conditions need utmost emphasis. 1.2 Research Objectives In this study, the primary goal is to develop performance-based pass/fail criterion for geosynthetic-reinforced asphalt concrete specimens. • Evaluate the effect of geosynthetic types, confining pressures, temperatures, and shear strain rates on geosynthetic-reinforced interlayer bonding performance. Thereby, develop the bond shear strength prediction models.
2 • Perform FlexPAVE™ analysis on various overlay pavement structures, speeds, temperatures, and overlay thicknesses to investigate the critical debonding condition for geosynthetic-reinforced specimens. • Determine the selection criterion for debonding resistant geosynthetic products. 1.3 Thesis Organization The thesis consists of five chapters. Chapter 1 introduces the research background, the objective of this study. Chapter 2 is the literature review on reflective cracking, numerous ways to mitigate reflective cracking, functions of geosynthetic, factors influencing the interlayer bonding performance, and bonding of geosynthetic-reinforced interlayer. Chapter 3 describes the materials, specimen fabrication process, and testing methodology, including Modified Advanced Asphalt Shear Tester and Digital Image Correlation system. Chapter 4 provides pavement response analysis on overlay structure pavement using FlexPAVE™. Chapter 5 describes interface shear strength (ISS) results, statistical analysis on tack coat application rates, shear strength prediction models, maximum shear ratio concept and simulation under various conditions, and proposing the selection criterion for geosynthetic products. Chapter 6 summarizes the conclusions and recommendations.
3 2. LITERATURE REVIEW
2.1 Reflective Cracking Reflective cracking is the common distress after overlay placed over the old cracked Portland cement concrete (PCC) or the hot mix asphalt (HMA) pavement. The existing crack on the underlaying old pavement causes the crack form at the bottom of the overlay and propagates through itself. Reflective cracking breaks the continuity of overlay and allows the water to enter the pavement. This will further reduce the soil-bearing capacity and deteriorate the entire pavement structure. Also, the prevalence of this distress would significantly influence the travel safety, driving comfortability, and service life of the pavement (Rigo 1993). Figure 2.1 shows the mechanism of reflective cracking. The temperature variation and repeated traffic loading can induce the stress concentration adjacent to discontinuities tip in the existing pavement. The initial crack forms and propagates through the overlay due to the bending, shear, and thermal contraction effects (Lytton 1989).
Figure 2.1 Mechanism of reflective cracking (Nithin et al. 2015). There are many ways to mitigate the reflective cracking, such as rubblization, milling, chip seal, sealing, increasing overlay thickness, and installing the stress absorbing membrane interlayer (SAMI) (Blankenship et al. 2004, Makowski et al. 2005, Zhiming 1997, Zhou and Sun 2000). Properly selected and constructed geosynthetic interlayer is one of the promising ways to
4 effectively mitigate or control the reflective cracking (Baek 2010, Khodaei and Falah 2009, Mukhtar and Dempsey 1996).
2.2 Function of Geosynthetic 2.2.1 Reinforcing Reinforcing function requires the geosynthetic products, such as paving fabric or paving grid, at the interface have significantly greater modulus (more than five times) than the asphalt mixture layer of which its embedded in (Lytton 1989). Sprague et al. (1998) also found the geosynthetic products with the stiffness greater than 200 kN/m at a strain between 2% to 5% develop sufficient reinforcement to overlay. When the reflective cracking reaches the reinforced- interlayer, the original perpendicular crack propagation will transfer to horizontal direction and travel below the reinforced-interlayer. The properly installed reinforcing-geosynthetic interlayer can indefinitely delay the prevalence of the reflective cracking (Button and Lytton 1987). The sufficient overlay thickness is also required to achieve the reinforcing function. The common thickness, recommended by the geosynthetic manufacture, to install geosynthetic interlayer is at least 3.81 cm (1.5 in) (Huesker 2015, TenCate 2019). 2.2.2 Stress Relieving Common stress-relieving geosynthetic has lower stiffness and stalls the reflective cracking at interlayer, though crack could still form at the top of the interlayer system and propagate through the overlay. The stress-relieving geosynthetic can store strain at low stress level and mitigate the reflective cracking (Lytton 1989, Sprague et al. 1998). 2.2.3 Water Barrier While the crack penetrates through the overlay, the geosynthetic products act as a barrier to prevent water infiltration and protect the underlying structure (Lytton 1989). The fully impregnated geosynthetic can significantly reduce the water permeability. However, extra care should be taken when compacting the overlay; the permeable overlay would allow more water to trap at the reinforced layer. This will cause the rapid failure of overlay because of moisture damage (Bognacki et al. 2007).
5 2.3 Debonding Problem The interlayer bonding between the asphalt surface and the underlying course is significantly influencing asphalt pavement performance (Khweir and Fordyce 2003, Kruntcheva et al. 2005). Different pavement layers act as a monolithic structure that efficiently transfer stress and strain caused by temperature change and repeated traffic loading. This requires adequate interlayer bonding. However, insufficient interlayer bonding leads to stress concentration and may result in debonding (Su et al. 2008). Debonding causes the slippage or delamination of the surface course; the premature distress significantly decreases the service life of the pavement (Hachiya et al. 1997, Peattie 1980, Sutanto 2009). Stress at interlayer caused by the moving traffic is shown in Figure 2.2. Raab and Partl (2004) found the tension mode, shear mode, or a mix of tension and shear mode phenomenon in fracture mechanic could characterize the debonding.
Figure 2.2 Stress at interlayer caused by moving traffic (Raab and Partl 2004).
2.4 Factors Influencing the Bonding 2.4.1 Tack Coat Type Asphalt emulsion is widely used in tack coat application in the field. Base on the emulsion curing time, it can be categorized into rapid setting (e.g., CRS-2), medium setting, and slow setting (e.g., SS-1, CSS-1) conditions. Based on the survey conducted by Mohammad et al. (2012), slow-setting emulsions are widely used in the world due to its easy to spray and low cost. The selection of the asphalt emulsion type highly depends on the construction window, traffic condition, and environment temperature. If it fails to meet the construction conditions, the interlayer bonding strength cannot be guaranteed. Thereby, leading to premature distress happens in the asphalt pavement. Asphalt emulsions are not commonly used in the geosynthetic- reinforced interlayer installation. Button and Lytton (2007) claimed the most emulsions have less
6 viscosity compared with asphalt binder, which may not provide enough bonding. Also, geosynthetic-reinforced interlayer requires high application rate for emulsions depends on its binder content. However, this will increase the curing time and difficulty in construction. Asphalt binder is one of tack coat material that could generate higher interlayer bond strength compared with the most asphalt emulsion. The application of the asphalt binder does not require curing time, thereby asphalt binder is recommended to use in geosynthetic-reinforced interlayer construction (Button and Lytton 2007). However, due to high viscosity compared with the asphalt emulsion, it requires to heat binder to a high temperature to spray evenly. Cutback asphalt should not be used for the polymeric type of geosynthetic because the solvent will remain in the geosynthetic layer and further deteriorate the polymer (Button and Lytton 2007). 2.4.2 Tack Coat Application Rate Tack coat application rate impacts the interlayer bonding performance. Excessive or lack of tack coat induces premature distress in the pavement. However, the current researchers have the debate over whether there is an optimum tack coat application rate (Al-Qadi et al. 2009, Bae et al. 2010, Mohammad et al. 2002, Raposeiras et al. 2013). Table 2.1 and Table 2.2 list the NCHRP report 712 and FHWA Tech Brief recommended tack coat application rate for different surface conditions, respectively (Mohammad et al. 2012). Table 2.1 NCHRP report 712 tack coat application rate (Mohammad et al. 2012). Residual application rate Surface type (gal/yd2) New asphalt mixture 0.035 Old asphalt mixture 0.055 Milled asphalt mixture 0.055 PCC 0.045
Table 2.2 FHWA tech brief tack coat application rate (Federal Highway Administration 2016). Residual application rate Surface type (gal/yd2) New asphalt mixture 0.02-0.05 Old asphalt mixture 0.04-0.07 Milled asphalt mixture 0.04-0.08 PCC 0.03-0.05
7 The asphalt retention rate of geosynthetic material should be taken into consideration when applying tack coat for the geosynthetic-reinforced interlayer. Amini (2005) suggested the practice tack coat application rate for geosynthetic is tack coat application rate for particular pavement surface type plus asphalt retention rate. However, excessive tack coat application may cause difficulty during geosynthetic installation (Button and Lytton 2007). 2.4.3 Curing Time There is a discrepancy in the effect of the curing time. WDOT found the curing time is not a significant factor in influencing shear strength (Tashman et al. 2006). However, it is commonly reported in the field that the uncured asphalt emulsion was lifted by the wheels of a haul truck, which failed to achieve the design tack coat application rate in the field. Trackless tack coat solves the problem of tracking. The setting time for trackless tack is from 5 to 15 min and a good bond strength can be provided (Bea 2010, Clark 2012, Mohammad 2011). 2.4.4 Surface Texture Wilson et al. (2016) found a higher interlayer bond strength for milled pavement; however, in their recent research, the milled specimens cored from the field did not show shear strength difference with the unmilled specimens. They claimed this phenomenon might occur due to the moisture damage to the milled specimens (Wilson et al. 2017). WDOT extracted the field core from both the unmilled and milled surface and found the milled surfaced texture provided a better shear resistance ability (Tashman et al. 2006). 2.5 Test Methods Various assessment methods in the field or laboratory shed light on interface shear properties. For laboratory assessment conditions, various tests can be performed on field-core or laboratory-prepared specimens. It should be noted that the laboratory test can accurately control the experiment setting and obtain better repeatability and reproducibility.
8
Figure 2.3 Fracture mechanics crack mode. Typically, using fracture mechanics interlayer bonding assessment tests are categorized as Mode I tension opening test, Mode II in-plane shear test, and Mode III out-of-plane shear test (Collop et al. 2011), as shown in Figure 2.3. The tests cover a wide range of methods and conditions to capture interface shear properties. Different test protocols correspond to various test equipment. Due to the multiple factors contributing to interface shear properties, the selection of the test method is closely related to the mode of loading, failure mode, and experiment accuracy. Mode II in-plane shear test mode is commonly used to characterize interface shear properties because it could test easily and highly mimic the in-situ condition, which is helpful to understand the mechanism of interface shear properties. In the Mode II in- plane shear test, it can further divide into a direct shear test and simple shear test. The interface shear property is controlled by various factors such as test temperatures, loading rates, material types, tack coat application rates, and interaction among those factors (Boulangé and Sterczynskia 2012). The prototype of the pavement interface shear test was established on the soil mechanics principle, Leutner shear test was developed in Germany and its counterpart was built in the U.S. by Uzan (1978). Mode II in the plane shear mode test can categorize into guillotine type direct shear test or shear box type simple shear test. Figure 2.4 shows the stress distribution for the direct shear test and simple shear test. It should note that the direct shear tests have significant shear stress concentration and the simple shear tests have a parabolic shear stress distribution.
9
(a) (b)
Figure 2.4 Shear stress distribution (a) Direct Shear Test; (b) Simple Shear Test (Raab et al. 2009). Generally, the direct shear test lacks the confining pressure apparatus, in order to arrest precise interface shear properties normal stress plays a critical role in dictating the interface asphalt mixture interlock and friction behavior. Therefore, some research institutes use an extra load cell or an actuator to induce normal stress. The shear test devices usually are installed in the servo-hydraulic loading system (e.g., Material Testing System/MTS), which controls the mode of loading, with an extra environment chamber to maintain the testing temperature. The typical laboratory-prepared specimen is a double-layer cylindrical or cubical shape. In order to solve the interface alignment issue, a gap is introduced between the shear device’s shearing ring. 2.5.1 Ancona Shear Testing Research and Analysis (ASTRA) Device The Ancona Shear Testing Research and Analysis (ASTRA) Device was developed in Italy in 2005 by Universita Politecnica delle Marche in Ancona, as shown in Figure 2.5. ASTRA is a simple shear tester and can perform shear tests on double-layer specimens. This shear box bears the cylindrical specimen with a 95 mm diameter. During the test, a constant vertical normal load is maintained on the specimen. Also, a linear variable displacement transducer (LVDT) is used in the ASTRA system to record the specimen deformation. The measuring system of the ASTRA records the interface shear stress and vertical direction displacement. Conducting ASTRA test at different deformation rates and temperatures can obtain the adhesion and friction parameters for constructing the Mohr-Coulomb type envelope (Pasquini et al. 2015).
10
Figure 2.5 Ancona Shear Testing Research and Analysis (ASTRA) Device (Pasquini et al. 2015). 2.5.2 Louisiana Interlayer Shear Strength Tester (LISST) Louisiana interlayer shear strength tester is developed by Louisiana state university, as shown in Figure 2.6. The test could perform on 100 mm or 150 mm diameter double-layer specimens. The tester is composed of a shearing frame and reaction frame; during the test, the shearing frame is connected to the loading system while the reaction frame remains stationary. This test controls the displacement rate of 2.54 mm/min and test temperature of 25 1C, the normal confining pressure is capable of applying up to 206.84 kPa (30 psi) (Mohammad et al. 2018).
Figure 2.6 Louisiana interlayer shear strength tester (LISST) (Mohammad et al. 2018). 2.5.3 Sapienza Direct Shear Testing Machine (SDSTM) Sapienza Direct Shear Testing Machine is developed by the research team in Sapienza University of Roma (Tozzo et al. 2014), as shown in Figure 2.7. The tests perform on 100 mm
11 diameter double-layer specimen. The gap between the two molds is 10 mm. This test controls the load with the frequency of 5 Hz and the testing temperature is 21 1C. The study found normal pressure plays an important role in interface fatigue property. The monotonic shear tests have the same trend on the influence of the normal pressure with the cyclic fatigue shear test.
Figure 2.7 Sapienza Direct Shear Testing Machine (SDSTM) (Tozzo et al. 2014). 2.5.4 Advanced Shear Tester (AST) The advance shear tester (AST) was designed in 2015 by Zofka et al. (2015), as shown in Figure 10. The shear test devices could be installed in a servo-hydraulic loading system with an extra environment chamber. The AST laboratory-prepared specimen is a double-layer 150 mm diameter cylindrical shape.
Figure 2.8 Advanced Shear Tester (AST) (Zofka et al. 2015). Zofka stated that the boundary condition for the shear test can be divided into the constant normal load, constant normal stiffness, and constant volume. Though the constant normal load condition is used by most of the shear devices, he proposed the constant normal stiffness condition has its edge. Constant normal stiffness condition mimics the low speed heavy track on the thin layer condition. Also, the dilation property at interlayer cannot be
12 comprehensively explained by constant normal load condition. Therefore, constant normal stiffness could be a suitable candidate for constant normal load condition to use in the shear device. Also, using a constant normal load should install a vertical actuator to maintain the load. However, the constant normal stiffness shear device uses die springs to maintain the confining pressure, which largely decreases the cost of the device. 2.5.5 Modified Asphalt Shear Tester (MAST) The North Carolina State University (NCSU) research team developed the modified asphalt shear tester (MAST) shown in Figure 2.9 by modifying AST (Cho 2016). MAST is capable of conducting the shear tests in both monotonic and fatigue modes of loading under confining pressure. The cylindrical specimen with a diameter of 101.6 mm (4 in.) extracted from a 152.4 mm (6 in.) gyratory samples or square-shape specimen with widths of 152.4 mm (6 in.) and 101.6 mm (4 in.) trimmed from a slab sample can be used for the MAST shear tests, whereas AST allows to use only cylindrical gyratory specimens of 150 mm (6 in.) in diameter. It is a well-known fact that air void gradient exists along the periphery of gyratory specimens (Chehab et al. 2000), thereby using a 101.6 mm (4 in.) specimen cored from a 150 mm (6 in.) gyratory sample in MAST nullifies the uncertainties of air void effect on the test outcomes. The initial confining pressure is controlled by the in-line load cell through tightening the bolts on the side panel. The technique employed to apply confining pressure remains same for both AST and MAST devices. However, approaches of fastening the specimens to both machines vary vastly. AST allows the user to tie the specimen directly to the device’s upper and lower jaw of the moving and stationary collar by tightening the threaded bolt and nut arrangement. The specimens during shear testing typically expand due to the aggregate rearrangement along the interface. Thereby, the frictional forces between the collar walls and the specimen could not hold the specimen in place that leads to slippage. This influences the shear and the confining load cell readings during the test. MAST test device addresses this issue by gluing the specimen firmly to a four-set shoe arrangement. The shoe is fastened to a stationary portion of the jig that is free to move horizontally (along the confining pressure load cell) with the aid of linear tracks. Thus, MAST device allows the free expansion of specimen during testing without any slippage along the walls of the shoe. Figure 2.10 shows the typical confining pressure during a MAST monotonic shear test. The confining pressure recorded from the in-line load cell varies from initial stress by 5% and stabilizes after reaching peak shear. The gap
13 between the fixed and movable side platens is 8 mm. Even though the specimen is fastened using shoes to the jig firmly, the large bending moment generated during the test causes rocking motion to the shoe. MAST device has the provision to monitor the on-specimen displacements during such events with the aid of non-contact digital image correlation (DIC) technique. MAST is designed to have the opening on one side, which allows the DIC system to track on-specimen displacement. All the aforementioned factors make MAST a superior device over AST.
Figure 2.9 Modified asphalt shear tester (MAST) (Cho et al. 2017b).
1000 Shear Stress 800 Confining Pressure
600
400 Stress(kPa)
200
0 0 20 40 60 80 Time (s) Figure 2.10 Typical MAST test result. Cho and Kim (2016) verified the time-temperature superposition principle on shear failure of double-layered asphalt concrete specimens with different tack coats and GlasGrid
14 interlayer. A shear strength prediction model was proposed to predict shear strength at various confining pressures, temperatures, and shear strain rates. Furthermore, FlexPAVE™ analysis was conducted to determine the potential debonding state. Shear ratio (SR), as shown in Figure 2.11, was calculated by the ratio of the FlexPAVE™ computed shear stress to the model predicted shear bond strength. The maximum shear ratio was presented as an index parameter to determine the pavement debonding potential (Cho et al. 2017a).
Figure 2.11 Shear ratio concept. 2.6 Bonding of Geosynthetic-Reinforced Interlayer Baek (2010) found the interface shear bond strength is a good indicator to reflect the potential of reflective cracking. The lower interface bond strength could increase the possibility of developing reflective cracking. The geosynthetic reinforcement at the interlayer could decrease the interlayer bonding (Canestrari et al. 2016, Pasquini et al. 2013). The Adequate bonding in geosynthetic-reinforced interlayer sufficiently distributes the stress and guarantees the functionality of geosynthetic-reinforced interlayer. The improper installation of the geosynthetic could not stall or mitigate the reflective cracking propagation, which will influence the durability of the pavement (Ferrotti et al. 2012, Vanelstraete and De Bondt 1997). Canestrari et al. (2006) used ASTRA to conduct the shear test on two types of glass geogrids (GG), polyester geogrid (PG), and geomembrane (GM) reinforced specimens. The reinforced the double-layer system controlled top layer as dense graded mix; the bottom layer was either dense graded mix or open grade mix. The result shows a higher mesh dimension (25 ×25 mm2) of the paving grid has better shear resistance ability than lower mesh dimension product (12.5×12.5 mm2). With the decrease of the mesh size, the residual friction angle from
15 friction envelope also reduces. The research shows the bottom layer surface condition does not have an impact on the shear strength of GM reinforced specimens. In another paper, Canestrari et al. (2016) found geosynthetic with a higher thickness could significantly decrease the interlayer shear strength. Vismara et al. (2012) conducted monotonic shear tests to investigate the performance of the geosynthetic-reinforced interlayer. Polypropylene nonwoven and fiberglass grid composite reinforced slab specimens were subjected to the Leutner shear tester at 5C and 25C with a constant deformation rate at 0.85 mm/s (2 in./min). An average 70% reduction in shear strength compared with the control specimen is found in geosynthetic-reinforced specimens. Ferrotti et al. (2011) performed the monotonic shear test on both paving grid reinforced and unreinforced specimens at 20C. During the tests three confining pressures (0, 0.2 and 0.4 MPa) were selected at the constant displacement rate of 2.5 mm/min (0.1 in./min). The shear strength and friction envelopes were obtained by the shear tests. Under unconfinement conditions, the shear strength decreases with the presence of the paving grid, however with the increase of the confining pressure, the trend reverses. Between grid reinforced specimen, the polymer-modified emulsion specimens show a higher shear resistance performance than conventional emulsion type under all three confining pressures. He found a similar residual friction angle for both grid reinforced and unreinforced specimens. He reported due to the poor interlayer bonding, one condition of slab compactor compacted double-layer specimen separated during the specimen coring process, this type of the specimen did not apply any tack coat. The author claimed this problem was contributing to the poor quality when asphalt loose mix fabricated in mix plant. Also, the same reason explained the unreinforced specimen shows higher variability in monotonic shear tests. Nithin et al. (2018) using Leutner shear tester to conduct the shear test on three different (Jute, Coir, and Synthetic GlasGrid) geosynthetic materials. He found with the presence of the geosynthetic the interlayer shear strength has a certain reduction. The reduction is in conjunction with the tensile modulus of corresponding geosynthetic products, the geosynthetic with higher modulus results in experiencing less shear strength reduction. The shear tests are reported to have more variability at lower temperatures and high deformation rates. When the temperature increases from 10C to 30C the shear strength decreases by nearly 80%.
16 2.7 Critical Summary There are many ways to mitigate the reflective cracking, such as rubblization, milling, chip seal, sealing, increasing overlay thickness, and installing stress absorbing membrane interlayer (SAMI). Geosynthetic interlayer is one of the promising ways to effectively mitigate or control the reflective cracking. The primary functions of the geosynthetic are reinforcing, stress-relieving, and waterproofing. The reinforcing function requires the geosynthetic material to have significantly greater modulus than the surrounded asphalt layer. It could redirect crack propagation at interlayer, which can indefinitely delay or mitigate reflective cracking. Stress-relieving geosynthetic products have lower stiffness and can store strain at low stress level. The fully impregnated geosynthetic could significantly reduce the water permeability. The proper installation, controlling overlay thickness, and overseeing compaction quality are also required to achieve the functions of geosynthetic. Tack coat is required to use in geosynthetic-reinforced interlayer construction. Cutback asphalt should not be used for the polymeric type of geosynthetic because the solvent will remain in the geosynthetic layer and further deteriorate the polymer. Geosynthetic-reinforced interlayer requires high application rate for emulsions depends on its binder content. However, this will increase the curing time and difficulty in construction. The application of the asphalt binder does not require curing time and application rate is satisfied for the construction requirement. Thereby, asphalt binder is recommended to use in the geosynthetic-reinforced interlayer. Some researchers suggest the practice tack coat application rate for geosynthetic is tack coat application rate for a pavement surface type plus asphalt retention rate. However, excessive tack coat application may cause difficulty during geosynthetic installation. The direct shear test is helpful in understanding the mechanism of interface shear properties. The geosynthetic material installing at the interlayer decreases the interlayer shear bonding. The improper installation of the geosynthetic could not stall or mitigate the reflective cracking propagation. The lower interface bond strength could increase the possibility of developing reflective cracking. Geosynthetic with a higher thickness significantly decreases the interlayer shear strength.
17 3. EXPERIMENTAL PLAN
3.1 Materials 3.1.1 Asphalt Mixture In this study, RS9.5C mixture with 40% RAP (hereinafter referred as RAP-40 mixture) was selected. Material characterization and verification of reported parameters in the job mix formula (JMF) were considered as the initial step before the commencement of performance testing. The virgin binder used for the RAP-40 mixture is PG 58-22. The total binder content in this mixture is 6.0%. Figure 3.1 presents the aggregate gradation of RAP-40. Considering the high RAP content in RAP-40, the compaction temperature was selected as 145C.
100
80
60
40 % % Passing
20 RAP-40 Gradation Control Points 0 0.0 1.0 2.0 3.0 4.0 Sieve Size (0.45 Power) (mm) Figure 3.1 Aggregate gradation of RAP-40 mixture. Even though the loose mix is collected from the HMA plant, a high probability of segregation of fine and coarse particles was expected while shoveling the AC mixes to the collecting buckets. Hence, a homogenization procedure was followed before fabricating any samples using the loose mix in the laboratory. Loose mix collection in plastic buckets was done as a first attempt. Henceforth, an add-on step was followed to the traditional separation procedure. The loose mix was removed from the plastic bucket as a single unit, as shown in Figure 3.2 (a). Thereafter the cloth bags were carefully removed with minimal loss of loose mix
18 (Figure 3.2 b). The obtained single AC mix lump is transferred to a metal bucket to follow the typical separation procedure (Figure 3.2 c).
Figure 3.2 (a) Removing AC loose mix with cloth bag from the bucket (b) Removing the AC lump from the cloth bag and (c) the loose mix lump inside the metal bucket. One batch in the separation procedure includes four of five gallon buckets with a total AC mix weighing more than 100 kg (200 pounds). The metal buckets with the loose mix are heated to a temperature that is 10C less than the compaction temperature for two hours. After that, one-fourth portion from each bucket was poured into four separation pans; each pan was further divided into 12 small boxes. This procedure aids to producing a well-mixed asphalt mix and storing the four buckets of the loose mix back in 12 cloth bags (three bags for each bucket). A storage bag constitutes of four small boxes, each randomly selected from four separation pans. Figure 3.3 shows the separation pans and cloth bags for storage. These separated mixes in the cloth bags will be used for sample fabrication and material characterization study. Depending on the material requirements, the separation process could be repeated.
19
(a) (b) Figure 3.3 Separation pans and cloth bags for making well mix AC. The theoretical specific gravity and the bulk specific gravity of the loose mix with
RAP40% were measured as per AASHTO T 209-12 (AASHTO 2012). It is found that Gmm for AC mixes with RAP40% is 2.44 g/cm3. Dynamic Modulus Test In order to prepare a cylindrical specimen for performance test with 6% air voids, four 38 mm cylindrical samples were cored from each gyratory sample with varying weight calculated based on AASHTO R83-17 (AASHTO 2017). The results of the air void study are shown in Figure 3.4. A linear relation is established between the weight of the gyratory sample and achieved air voids. This relation helps to predict the exact weight required for a 180 mm height gyratory sample to produce four - 38 mm diameter and 110 mm height cylindrical samples with 6% air voids from it. These cylindrical specimens with 6% air voids will be used to conduct the performance tests like dynamic modulus test and cyclic fatigue test to measure the material input parameters for the finite element (FE) software named as FlexPAVE™ version 1.1.
20
12.0 y = -0.02x + 114.75 10.0
8.0
6.0
4.0
2.0
Specimen Achieved Air Voids (%) Voids Air Achieved Specimen 0.0 6800 6900 7000 7100 7200 7300 Mass of Gyratory Sample (g)
Figure 3.4 The linear relation between the air voids and weight of gyratory samples. Asphalt concrete is thermorheologically simple within the linear viscoelastic range.
Therefore, by using a shift factor (aT), the effects of time and temperature can be represented by a unique parameter referred to as reduced frequency, fR. The linear viscoelastic properties of asphalt mixtures can be determined by dynamic modulus (|E*|) tests that measure a specimen’s stress-strain relationship under continuous sinusoidal loading. The parameters obtained are the complex modulus values and time-temperature (t-T) shift factors. Table 3.1 provides a summary of the linear viscoelastic properties of RAP-40 at selected frequencies and temperatures. Table 3.1 Linear viscoelastic properties of the test mixture. Test Conditions RAP-40 Temperature Frequency Complex Phase (C) (Hz) Modulus (MPa) Angle () 10 1.56E+04 8.32 4 1 1.23E+04 10.56 0.1 9.20E+03 13.60 10 7.87E+03 16.93 20 1 4.92E+03 22.17 0.1 2.74E+03 28.13 10 2.04E+03 32.84 40 1 8.84E+02 35.74 0.1 3.83E+02 34.99
21 Figure 3.5 presents the dynamic modulus test results for three replicates of the RAP-40 mixture at different temperature/frequency combinations. A mastercurve, a single smooth curve, was developed by shifting these data points horizontally at an arbitrarily selected reference temperature. The t-T shift factor is the amount of horizontal shift in the log scale that is required to create a continuous mastercurve. For thermorheologically simple materials, the amount of shifting is dependent only on the chosen reference temperature. Therefore, the shift factor varies by temperature, as shown in Figure 3.6 for the three replicates of the study mixture at a reference temperature of 20C (68F).
1.0E+05
1.0E+04
1.0E+03
1.0E+02 RAP-40_Rep 1
1.0E+01 RAP-40_Rep 2 Dynamic (MPa) Dynamic Modulus RAP-40_Rep 3 1.0E+00 1.0E-02 1.0E-01 1.0E+00 1.0E+01 1.0E+02 Frequency (Hz)
Figure 3.5 Dynamic modulus values for RAP-40.
22
3 y = 0.0006x2 - 0.155x + 2.855 2
1
) T
0 log (a -1
-2
-3 -10 0 10 20 30 40 50 Temperature (C)
Figure 3.6 Time-temperature shift factor function for RAP-40 mixture. Figure 3.6 presents the averaged t-T shift factor function coefficients for the mixtures, which were used later for the horizontal shifting of the interface shear bond strength data obtained by the Modified Advanced Shear Tester (MAST) at several different temperatures and constant displacement control rates. Figure 3.7 presents the dynamic modulus mastercurves for the RAP-40. Such mastercurves can be used to describe the constitutive behavior of asphalt concrete over a wide range of temperatures and frequencies.
23
1.0E+05
1.0E+04
1.0E+03
1.0E+02 RAP-40_Rep 1 RAP-40_Rep 2
Dynamic Dynamic Modulus(MPa) 1.0E+01 RAP-40_Rep 3 Sigmodial Fit 1.0E+00 1.0E-06 1.0E-03 1.0E+00 1.0E+03 1.0E+06 Reduced Frequency (Hz)
Figure 3.7 Dynamic modulus mastercurve for RAP-40. 3.1.2 Tack Coat Tack coat application Table 3.2 provides a summary of the tack coat application rates used in this project. The NCDOT QMS manual (North Carolina Department of Transportation (NCDOT)) 2018), Table 605-1, stipulates an optimal rate of 0.181 L/m2 (0.04 gal/yd2) for emulsified tack coats, which eventually leaves a residue of 0.03 gal/yd2. A few studies report that the variability in the target and achieved application rates of tack coats in the field ranges from 4% to 106% (Al-Qadi et al. 2008, Mohammad et al. 2012). Based on those outcomes, the dry and wet conditions in the field are mimicked by varying the residual application rate by 66% of the optimal residual application rate for unreinforced sections, i.e., ± 0.091 L/m2 (0.02 gal/yd2). Hence, for unreinforced specimens, three residual application rates of PG 64-22 asphalt binder, 0.045 L/m2 (0.01 gal/yd2), 0.136 L/m2 (0.03 gal/yd2), and 0.226 L/m2 (0.05 gal/yd2) were used. The optimal application rates used for the geosynthetic-reinforced specimens follow the manufacturers’ recommendations. The dry and wet application rate conditions were determined by adding and reducing ± 0.091 L/m2 (0.02 gal/yd2) from the manufacturer’s optimal application rate, respectively.
24 Table 3.2 Summary of tack coat application rates for geosynthetic products used in this study. Asphalt Application Rate Wet Optimal Dry Retention Rate L/m2 Units gal/yd2 Control Specimen Metric 0.23 0.14 0.05 - (Effective Tack Coat Application Rate) Imperial 0.05 0.03 0.01 - Metric 0.59 0.5 0.41 0.36 PC#1 HaTelit G50 (40 x 40) Imperial 0.13 0.11 0.09 0.08 Composite Metric 1.45 1.36 1.27 1.22 PC#2 Mirafi MPG100 Imperial 0.32 0.3 0.28 0.27 Paving Metric 0.77 0.68 0.59 0.54 PM GlasPave 50 Mat Imperial 0.17 0.15 0.13 0.12 Paving Metric 0.32 0.23 0.14 0.09 PaG Glasgrid 8511 (25) Grid Imperial 0.07 0.05 0.03 0.02 Paving Metric 1.13 1.04 0.95 0.91 PF PetroMat 4598 Fabric Imperial 0.25 0.23 0.21 0.20
Complex Shear Modulus (|G*|) The application of the time-temperature (t-T) superposition principle to determine the ISS and interface shear stiffness of geosynthetic-reinforced asphalt concrete has been verified by Cho and Kim (Cho and Kim 2016). In that study, GlasGrid-reinforced asphalt concrete core specimens were sheared in the Modified Asphalt Shear Tester (MAST) under constant displacement rate to measure the ISS at different test temperatures. The t-T shift factors (aT) of the asphalt concrete mixture that were measured via dynamic modulus testing were used initially to verify the t-T superposition principle for the MAST outcomes. The MAST test results obtained from unreinforced specimens were used to create a mastercurve with the aid of the mixture aT. However, the MAST test results for the geosynthetic-reinforced specimens demonstrated spurious results with the same mixture aT. Therefore, using a dynamic shear rheometer (DSR), frequency sweep tests were conducted using a tack coat asphalt binder to obtain the asphalt binder t-T shift factor (aT). The asphalt binder aT was used successfully to construct an ISS mastercurve for the geosynthetic-reinforced specimens. Thus, the conclusion drawn from the Cho et al. (2017) study is that the mixture aT applies to unreinforced MAST specimens, whereas the asphalt binder aT is applicable for geosynthetic-reinforced specimens.
25 In this study, ISS tests using the MAST were carried out on geosynthetic-reinforced specimens. Asphalt binder PG 64-22 was used as the tack coat for the project. As explained, the tack coat asphalt binder aT is a requirement for mastercurve construction. Thus, frequency sweep tests were conducted using the PG 64-22 binder to determine its t-T shift factor (aT). These tests were completed using an Anton Paar MCR 302 DSR with 8-mm parallel plate geometry. The frequency sweep test typically applies a load frequency that ranges from 0.1 Hz to 30 Hz at 1% shear strain amplitude. The selected test temperatures were 5C (41F), 20C (68F), 35C (95F), 50C (122F), and 64C (147.2F). Complex shear modulus (|G*|) mastercurves were constructed based on the Christenson–Anderson–Marasteanu (CAM) model (Christensen and
Anderson 1992), as given in Equation (1) where |G*|g is the glassy dynamic shear modulus and is equal to 1 GPa for asphalt binder, and ωc, m, and v are the CAM model fitting parameters for the
|G*| mastercurve. Equation (2) describes the reduced frequency, ωR, where aT is the shift factor at temperature T and ω is the actual testing frequency.
m − v v **c GG()1R =+ (1) g R
RT=a (2)
Equation (3) was employed to fit the t-T shift factor, where α1, α2, and α3 are the fitting coefficients, as shown in Table 3.3.
2 log aTTT =++123 (3) Table 3.3 Averaged time-temperature shift factor function coefficients for PG 64-22.
α1 α2 α3 0.000823 -0.15368 2.741548
Figure 3.8 presents the (a) binder dynamic shear modulus mastercurve fitted by the CAM model and (b) the t-T shift factor response of the PG 64-22 binder.
26
1.E+10
1.E+08 (Pa)
1.E+06
1.E+04
PG 64-22_Rep-1
Dynamic Dynamic Modulus 1.E+02 PG 64-22_Rep-2 Fit 1.E+00 1.00E-06 1.00E-03 1.00E+00 1.00E+03 1.00E+06 Reduced Frequency (Hz)
(a) 4 y = 0.001x2 - 0.168x + 3.083
2 )
T 0
log (a log -2
-4
-6 -20 0 20 40 60 80 Temperature (C)
(b) Figure 3.8 (a) Dynamic shear modulus mastercurve and (b) shift factor response of PG 64-22.
27 3.2 Specimen Fabrication 3.2.1 Compaction AFG2 Superpave gyratory compactor (Pine Test Equipment, Inc.), shown in Figure 3.9, was used to fabricate double-layered MAST samples. The separated loose mix in cloth bags was heated to the compaction temperature for an hour. Then, the loose mix was batched in required quantities and placed into pans depending on whether it would be fabricated as a top or bottom layer. The pans containing the loose mix were placed in an oven at the compaction temperature for another hour. The mold and necessary test accessories (such as spatulas) also were heated in the same oven for an hour to reach the compaction temperature. Subsequently, the bottom layer was compacted to a height of 50.8 mm (2 in.) in the gyratory mold that was 150 mm (6 in.) in diameter
(a) (b) Figure 3.9 (a) Superpave gyratory compactor (Pine Test Equipment, Inc.) and (b) compaction molds After the tack coat application process was completed, the bottom layer was placed back into the hot mold, and the top layer was compacted directly on top of it. No wait time was needed for the top layer compaction because the hot binder was used as the tack coat. The top layer material preparation procedure was the same as for the bottom layer. The height of the top layer was 50.8 mm (2 in.), making the total height of the MAST sample 101.6 mm (4 in.). Figure 3.10 presents the double-layered MAST sample compaction procedure.
28
150 mm 50.8mm
Bottom Layer
Geosynthetic (a)
Tack Coat 50.8mm
Bottom Layer
150 mm 50.8mm (b) Top Layer
Geosynthetic
Tack Coat 50.8mm
Bottom Layer
(c) Figure 3.10 Compaction procedure for double-layered MAST sample: (a) bottom layer fabrication, (b) bottom layer placement in hot mold with tack coat, and (c) completed MAST sample. 3.2.2 Geosynthetic Material Preparation The interface of the asphalt concrete layers in this study was constructed as either a tack coat only (control specimen) or a tack coat-impregnated geosynthetic interlayer. Five different types of geosynthetic reinforcements were used as interlayers in the current project. Table 3.4
29 provides the nomenclature details for each geosynthetic product, and the acronyms are used hereafter. Table 3.5 presents the properties of geosynthetic. Table 3.4 Nomenclature details of different types of geosynthetics. Acronym Acronym expansion Product Details PC#1 Paving Composite #1 Huesker HaTelit G50 PC#2 Paving Composite #1 TenCate Mirafi MPV 100 PM Paving Mat Tensar GlasPave 50 PF Paving Fabric Propex PetroMat 4598 PaG Paving Grid Tensar GlasGrid 8511
The geosynthetic interlayer installation procedure involves three stages: cutting the geosynthetic products, applying the tack coat, and placing the geosynthetic product. Details regarding each stage are given below. Figure 3.11 shows the five different types of geosynthetic products used for the current project.
(a) (b) (c)
(d) (e) Figure 3.11 Geosynthetic samples: (a) PC#1, (b) PC#2, (c) PaG, (d) PM, and (e) PF.
30 Table 3.5 Properties of geosynthetic products. Strip Tensile tensile Grab Tensile Mass/uni Melting Asphalt Physical Properties strength strength tensile elongatio t area point retention (kN/m) (N/50 strength n mm) Metric g/m2 kN/m N/50 mm N % C Lm Units Imperial oz/yd2 lbs/in. lb/2 in. lb % F gal/yd2 HaTelit G50 Metric 270 50 - - 3% 255 (Bitumen PC#1 Composit (40 x 40) Imperial 8 285 - - 3% 490 coated > 60%) e Mirafi Metric 678 115 - - 3% 800 1.2 PC#2 MPG100 Imperial 20 655 - - 3% 1472 0.27 Paving Metric 237 50 <5 >232 0.47 PM GlasPave 50 Mat Imperial 7 280 <5 >450 0.1 Metric 405 100 3% >232/>820 Pressure Paving GlasGrid sensitive PaG >450/>150 Grid 8511 (25) Imperial 12 571 adhesive 3% 8 backing Paving PetroMat Metric 139 - - 449 50% 160 0.91 PF Fabric 4598 Imperial 4.1 - - 101 50% 320 0.2
31 The interface of the asphalt concrete layers in this study utilized either a tack coat only or a tack coat with an interlayer. Different types of geosynthetic reinforcements were used as the interlayers here. The geosynthetic interlayer installation procedure involves three stages: (1) cutting the geosynthetic products, (2) applying the tack coat, and (3) placing the geosynthetic. The details of each stage are explained in the sections below. 3.2.3 Interface Installation Cutting the geosynthetic products In order to fabricate the geosynthetic-reinforced MAST samples using a gyratory compactor, geosynthetic circular samples 140 mm (5.51 in.) in diameter were cut from the roll of geosynthetic material. The machine direction (MD) of geosynthetic products typically requires that the products are unrolled in the traffic direction, as shown in Figure 3.12. The geosynthetic samples are extracted in the diagonal direction to avoid any manufacturing defects in the MD and cross-machine direction. Figure 3.13 (a) illustrates the template used for the trimming pattern. Figure 3.13 (b) and (c) show the template traced over the PC#1 and PM, respectively.
Machine Direction (MD)
Cross Cross MD Traffic Direction
Figure 3.12 Placement of geosynthetics in the field.
32
Composite HaTelit G50
(b) diagonal Paving Mat GlasPave 50
140 mm
(a) (c) Figure 3.13 Geosynthetic interlayer sample cutting process. A cloth cutter (Reliable 1500 FR) that can cut up to 1-inch thick fabric bundles was used here to cut the geosynthetic sample shapes from the rolls. The first step is to draw the outlines using a template and a peel-off marker, as shown in Figure 3.14. The cutter is run through the trace mark to extract the geosynthetic sample.
(a) (b) (c) Figure 3.14 (a) Tracing the cutting pattern, (b) completed template pattern, and (c) cutting/extracting the geosynthetic sample using a cloth cutter.
33 Geosynthetic placement Figure 3.15 shows the placement of four of the geosynthetic products.
(a) (b) (c)
(d) (e) Figure 3.15 Placement of Geosynthetic Interlayers (view towards bottom layer): (a) PC#1, (b) PC#2, (c) PaG, (d) PM, and (e) PF. The placement technique that is used for each geosynthetic product is based on the manufacturer’s guidelines. The PaG is made of glass fibers, with one side having self-adhesive properties. The manufacturer stipulates that the adhesive side of the grid facing the bottom layer is placed first, and then the specified rate of tack coat is applied onto the grid afterward. However, other geosynthetic products are installed after the tack coat application. For the installation of the two geosynthetic composites, the PC#1 should be placed with the grid side in contact with the top layer. PC#2 should be placed over the bottom layer with the grid side facing it. The PM has two different colors on either side. A thin layer of asphalt applied to one side makes that side black, whereas the side without the tack coat is grey. After applying the optimal application rate, the paving mat is placed with its grey side touching the tack coat. Note that, before placing the geosynthetics, the bottom layer is heated to 145C for two minutes to liquify the asphalt binder so that proper impregnation of the asphalt and gluing are achievable.
34 Following geosynthetic placement, a set pressure is applied by rolling a metallic rod over the specimen to ensure proper bonding throughout the contact area between the product and bottom layer, as shown in Figure 3.16. Figure 3.15 shows the top view of four different geosynthetic products on the bottom layer.
(a) (b)
(c) Figure 3.16 Setting pressure application on PC#1 using metal rod: (a) bottom layer after tack coat application, (b) liquified asphalt binder after placing the bottom layer with tack coat in the oven at 145C for two (2) minutes, and (c) the setting pressure application over the PC#1 by rolling the metallic rod.
35 The direction of the geosynthetic installation is a critical factor that determines the shear performance of the specimen. The MD of the geosynthetic is labeled on the samples throughout the sample fabrication process, as shown in Figure 3.17.
(a) (b) Figure 3.17 Tracking the geosynthetic-reinforcement placement direction (a) bottom layer of the sample (b) final cored specimen. 3.2.4 Tack Coat Application A day-long cooling period was allowed prior to the application of the tack coat, which in this case was hot binder PG 64-22 that was applied to the specimen to provide the desired interface bond. For this purpose, a metal canister with small holes in its cap was used to pour the hot binder to create a dense binder grid on top of the bottom layer, as shown in Figure 3.18(a). Then, a heat gun and metal spatula were used to warm and then spread the binder evenly on the surface of the specimen, as shown in Figure 3.18(b). However, the optimal application rate stipulated by the NCDOT (0.04 gal/yd2 for emulsified asphalt, Table 605-1) created challenges in applying the tack coat uniformly above the bottom layers. Figure 3.19 presents the specimen surfaces with different tack coat application rates, which show the non-uniform application of tack coat, especially at low application rates.
36
(a) (b) Figure 3.18 Tack coat application process: (a) pouring hot binder from the metal canister with perforated lid and (b) spreading binder uniformly using a heat gun and the metal spatula. Subsequently, laboratory equipment that could be used to apply the tack coat in a uniform pattern above the AC surface is searched. This search resulted in procuring a hot spray gun to apply hot molten glue. The following section depicts the procedure for achieving a uniform tack coat using a hot spray gun.
Figure 3.19 Non-uniformity found on tack coat applied bottom layer surface of MAST samples Tack coat application using a hot spray gun for MAST samples Figure 3.20 shows the test set-up for using a hot spray gun to apply the tack coat. The hot spray gun dispenses hot liquids under pressure. In the current study, hot asphalt is sprayed. The system consists of a hot liquid applicator and a control panel with a stand that holds the
37 applicator gun when it is not in use. Compressed air control and manometer indication are integrated into the control panel.
Air inlet Control Panel
Air spray regulator Liquid flow regulator
(b)
Spray trigger Hot Spray gun
Temperature controller
(a) (c) Figure 3.20 Details of: (a) test setup (b) the control panel (c) the hot spray gun.
38
Air spray nozzle
(b)
Liquid nozzle
Cartridge
(c)
(a) Figure 3.21 Hot spray gun: (a) components, (b) air spray nozzle, and (c) liquid nozzle. The step by step procedure for using the spray gun is as follows and illustrated in Figure 3.22. Step 1: Heat the asphalt binder can and cartridge (without nozzle) to 145C for 1 hour in the oven. Meanwhile, set the temperature of the spray gun (without cartridge) for 20 minutes at 175C (350F) to preheat the heating chamber. Step 2: Attach the liquid nozzle to the cartridge after one hour; Figure 3.22 (a). Step 3: Pour the asphalt to ¾ capacity (roughly 250 ml); Figure 3.22 (b). Step 4: Load the cartridge into the heating chamber; Figure 3.22 (c). Step 5: Close the lid to pressurize the asphalt cartridge; Figure 3.22(d). Step 6: Measure the temperature at a location shown in Figure 3.22 (f) for an accurate reading because the temperature inside the heating chamber is not calibrated. The trial study recommends the set temperature to be maintained at 175C (350F). Otherwise, for a set temperature of 145C (293F), the maximum achievable temperature is 120C. An infrared thermal gun could be used to measure the temperature, as shown in Figure 3.22 (e).
39 Step 7: Begin the spraying process once the spray applicator reaches the required temperature. Step 8: Allow sufficient heating time between each spraying sequence, and a proper temperature check should be done prior to the application. Step 9: Clean the spray applicator and control panel using a cloth moistened with a citrus solvent blended with hyper surfactants, commercially known as ‘orange cleaner.’ Meanwhile, the cartridge, liquid, and air nozzle must be cleaned using a kerosene-based solvent.
(a) (b) (c)
(d) (e) (f) Figure 3.22 (a) Attaching liquid nozzle to hot cartridge, (b) pouring liquid asphalt into cartridge, (c) loading cartridge into heating chamber of gun, (d) cartridge with asphalt inside spray gun, (e) closing spray gun mouth, and (d) measuring the temperature at nozzle tip.
40 With regard to the application of the hot binder (PG 64-22), a few trial tests were carried out on white paper to measure the optimal air pressure and liquid flow for PG 64-22. The results indicate that, as the air spraying pressure increased, the atomization of the asphalt became evident. However, increasing the liquid flow pressure above 1 bar could result in excessive flow. Therefore, based on this study, the air spraying pressure should be maintained between 4 to 5 bars while the liquid flow pressure is below 1 bar. Also, sufficient heating time between applications (5 to 10 minutes) should be provided to allow proper atomization of the asphalt.
Bottom layer of gyratory samples Gap between sample and plate
Opening to read the weighing scale
(a) (b)
(c) Figure 3.23 (a) Sheet cover above gyratory sample, (b) small gap between the cover sheet and sample, and (c) application of hot asphalt using a hot spray gun. A metal cover sheet with standing legs was designed in such a way that none of the excess tack coat spray could reach the sample side walls or scale. The test set up is shown in Figure 3.23 (a). The gyratory cover sheet diameter is 140 mm (5.5 in.). The metal sheet should not touch the sample because the weight measured by the scale should be only the tack coat
41 applied above the 140-mm diameter circular area, as shown in Figure 3.23 (b). The application rate was recalculated for the metal cover sheet opening, i.e., 140 mm. Figure 3.24 shows apparent differences between the surfaces where the tack coat was applied using the metal canister versus the hot spray gun. The expected sample-to-sample variation in the case of the hot spray gun is 0.3 g, but the tack coat will be spread uniformly across the surface. The tack coat weight was checked at every application cycle. The definition of ‘application cycle’ is a single pass of moving the hot spray gun from left to right and vice versa continuously from top to bottom, then returning to the top where the application commenced. This process ensures uniformity above the surface.
0.03 gal/yd2 2.2 g
(a) (b) Figure 3.24 Tack coat applied to the bottom layer surface of MAST samples: (a) non-uniformity (metal canister) and (b) uniformity (hot spray gun). 3.2.5 Sample to Specimen Fabrication In order to eradicate the effect of an air void gradient and to make uniform and consistent test specimens, the gyratory-compacted samples were cored and trimmed to the final specimen dimensions of 101.6 mm (4 in.) in diameter and 76.2 mm (3 in.) in height. Figure 3.25 presents the coring and cutting procedure. Specimen cutting As shown in Figure 3.26, a PVC pipe 100 mm in diameter was used to protect the specimen during cutting. The pipe serves to hold the layers together and absorb the vibrations and bending forces imparted by the saw, thereby protecting the specimen from damaging the weak interfaces. In addition, the PVC pipe can be used as a guide scale to fix the trimming location. First, as shown in Figure 3.26 (a), the sample is placed on a metal canister to keep the
42 sample level. Figure 3.26 (b) shows that the PVC pipe with leveling marks is matched with the bottom of the specimen, i.e., at the top of the canister. Figure 3.26 (c) shows the completed geosynthetic-reinforced sample with PVC pipe protection. This procedure saves time and reduces human error by producing highly repeatable specimens (in terms of dimensions) compared to the typical pen-marking method. Figure 3.27 shows the side view of geosynthetic- reinforced specimens with four different products.
100 mm 38.1 mm
38.1 mm (a)
Coring & Cutting
50.8mm
(b) 50.8mm
150mm
(c) Figure 3.25 Procedure for coring and cutting cylindrical specimens: (a) cored MAST sample, (b) trimming the top/bottom layer, and (c) finished specimens.
43
(a) (b) (c) Figure 3.26 Process for PVC pipe protection of geosynthetic-reinforced sample.
(a) (b) (c)
(d) (e) Figure 3.27 Side view of geosynthetic-reinforced specimens with: (a) PC#1, (b) PC#2, (c) PaG, (d) PM, and (e) PF.
44 3.2.6 Air Void Study on MAST Specimen The production of consistent MAST specimens plays a crucial role in interface shear tests. In order to achieve uniform samples, air void content in the gyratory-compacted specimens is investigated. The major aim of the air void study was to obtain the same air void content for both the bottom and top layers. Table 3.6 presents the air void contents of the MAST specimens obtained using the saturated surface-dry method. A clear difference in the achieved air voids is evident when the target air void content of the same sample was used for the top and bottom layers. Moreover, when targeting an air void content below 6% for the bottom layer, the gyratory compaction number goes above 130, which is far above the Ndesign of 75 gyrations. The excessive shear causes the aggregate to break, thereby negatively affecting the asphalt mixture performance and the interface bond. Hence, in this study, the air void content of 7% was targeted for the MAST specimens. Table 3.7 presents further air void content verification. Table 3.6 MAST specimens: air void content. No. 1 2 3 4 Sample Target Air Void Content 8.7% 8.2% 7.8% 7.3% Bottom 73 94 139 140 No. of Gyrations Top 21 27 29 35 Sample 8.1% 7.5% 7.0% 6.1% Achieved Air Void Specimen 7.3% 6.6% 6.4% 5.5% Content Top Layer 7.4% 7.2% 7.0% 6.3% Bottom Layer 6.9% 6.1% 5.9% 5.0%
Table 3.7 Verification of air void study results. Sample Target Air No. of Gyrations Achieved Air Void Content Void Content Top Bottom Bottom Top Specimen Top Layer Bottom Layer 8.70% 8.00% 70 22 6.94% 7.15% 6.76%
Figure 3.28 shows the relationship between the sample target air void contents and the achieved air void contents for the different beam specimen layers. Based on these results, the design air void contents of 8.7% and 8.0% were chosen for the top and bottom layer compaction, respectively.
45
8.0% No. of Gyrations y = 0.8105x + 0.0045 21 7.0% 27 29 73
6.0% 35 94
139 5.0% y = 1.3058x - 0.0448 RAP-40 Top 140
Specimen Achieved Air Air Voids Achieved Specimen RAP-40 Bottom 4.0% 7.00% 7.50% 8.00% 8.50% 9.00% Sample Target Air Voids
Figure 3.28 Target and achieved air void content relationship for different layers.
3.3 Testing Methodology 3.3.1 Modified Asphalt Shear Tester Recent work at NCSU has employed a shear device, known as the Modified Advanced Shear Tester (MAST), and shown in Figure 3.29, to perform confined shear tests on layered asphalt specimens, both with and without geosynthetic interlayer systems. Most notably, this work allowed the development of shear strength mastercurves for bond strength values at various combinations of loading rate, temperature, tack coat material, and confining stress.
46
Figure 3.29 Illustrations of the Modified Advanced Shear Tester (MAST).
Gluing the specimen to MAST shoes The MAST shear test is a quick test that induces substantial loading in a short time at the specimen interface. The boundary conditions of the MAST test set-up develop bending moments within each layer of the specimen that rotates the specimen. However, a weak grip between the specimen and the testing jig could lead to slippage in the normal/confinement direction and lead to erroneous outcomes. Therefore, utmost care was taken in this study to affix the specimen to the jig with the aid of metallic ‘shoes.’ A MAST shoe set consists of two pairs of shoes and a shoe frame that holds the specimen in the shoe for gluing using epoxy. An 8-mm gap is present between the shoes in each pair. The step-by-step procedure for gluing the specimen to the jig is explained below and shown in Figure 3.30.
47
(a) (b) (c)
(d) (e) (f) Figure 3.30 Gluing procedure for MAST specimen: (a) tightened bottom shoes on the gluing jig (b) application of glue on the bottom shoes, (c) specimen placement on the bottom shoes, (d) the upper shoe installation above the specimen, (e) specimen with all shoe in place, and (f) trimming the extra glue from the shoe edges. 1. In accordance with the manufacturer's recommendation for the preparation of the epoxy glue, mix the glue agents at a ratio of 9:1, i.e., 54 g of plastic steel putty (black) and 9 g of putty hardener (white) for a pair of shoes. Place the necessary quantity of the glue agents on a mixing cardboard card and mix well for two minutes to achieve a uniform grey paste.
48 2. Place the bottom shoe on the metal frame and affix it to the frame using screws, as shown in Figure 3.30 (a). 3. Divide the glue on the mixing card into four equal portions and apply two portions. Repeat for the second pair of shoes. 4. Carefully place the specimen on the bottom shoe, as shown in Figure 3.30 (c). Position the interface of the specimen such that it is at the center of the gap between the shoes. Note: The geosynthetic installation direction marked on the specimen needs to be in line with the shearing direction, i.e., perpendicular to the ground. 5. Place the upper shoe over the specimen and tighten it inside the frame. Fill any gap between the lower and upper shoes with metal plates. 6. Remove any excess epoxy glue that squeezes out at the specimen edges using a spatula. 7. Allow a glue curing period of 16 hours prior to using the MAST.
Specimen glued to MAST jig Once the MAST specimen is glued to the shoes and the speckled papers are pasted onto the shoes, the shoe must be loaded to the MAST testing jig, as shown in Figure 3.31 (a). The horizontal translation of the MAST shoe is constrained by installing a collar that ties the shoe to the jig. Figure 3.31 (b) shows the confining plate with a load cell added to the jig. Figure 3.31 (c) shows the MAST placed over the material test system (MTS – 810). Once the necessary connection between the actuator piston and the MAST vertical loading rod is established, then the environmental chamber is installed for conditioning the specimen to the test temperature, as shown in Figure 3.31 (d). Figure 3.31 (e) shows the digital image correlation (DIC) test set-up, and Figure 3.31 (f) shows the view through the charge-coupled camera. After three hours of conditioning, the MAST test is commenced.
49
(a) (b) (c)
(d) (e) (f) Figure 3.31 (a) Loading the MAST shoes with the specimen to the MAST jig, (b) installing the confining pressure plate with load cell, (c) placing the MAST over the MTS 810, (d) environmental chamber, (e) DIC test set-up, and (f) view through DIC camera.
MAST test configuration The geosynthetic-reinforced specimens were subjected to monotonic shear tests once the specimens reached the test temperature after three hours of conditioning. A load that corresponds to the normal stress was applied by tightening the confinement plate against the specimen, as shown in Figure 3.32. All the geosynthetic-reinforced specimens were sheared at a constant displacement rate (crosshead) of 5.08 mm/min (0.2 in./min). Throughout the test, the shear force, the normal vertical stress, and the horizontal displacement were continuously recorded. In
50 addition, the Digital Image Correlation (DIC) technique was used to measure on-specimen displacement. Figure 3.33 shows the DIC settings with regard to the MAST.
Figure 3.32 Loading configuration of MAST test set-up.
Figure 3.33 DIC settings on the MAST to monitor on-specimen displacement. 3.3.2 Digital Image Correlation The digital image correlation (DIC) technique was employed to measure on-specimen displacements. For this purpose, a sequence of images was captured at regular intervals of 150 ms during the test. The DIC algorithm needs a distinct random point to track the movement of the specimen between consecutive images and determine the displacement, as shown in Figure 3.24.
51
xxuxypp* =+( , )
x yyvxypp* =+( , ) MAST Shoe
Gray scale from 0 to 100 yp* 5 25 15 65 y p 100 50 25 95
35 40 15 35
80 0 25 95 x x p p* y Before Deformation After Deformation
Figure 3.34 Digital image correlation process (Safavizadeh and Kim 2017). Speckled paper is used for this purpose. Bits of speckled paper were cut to dimensions of 3.75 cm by 2.45 cm. The samples were placed in a tray and sprayed with matte black paint from a specified distance, as shown in Figure 3.35 (a). Figure 3.35 (b) shows the finished speckled paper. The other side of the speckled paper was sprayed with glue and pasted onto the MAST shoe, as shown in Figure 3.35 (c).
(a) (b) (c) Figure 3.35 Preparation of speckled paper: (a) spray painting the paper, (b) finished speckled paper, and (c) speckled paper on MAST shoe to track on-specimen displacement. Calibration of the DIC System for MAST Testing Even though the principle behind a monotonic asphalt shear test is simple, the MAST is a complex device. It has numerous components that must be assembled and disassembled during specimen loading and unloading.
52 The major challenge during MAST testing is to ensure that shearing occurs only along with the interface, which is achieved by securely attaching shoes to the jig through threaded bolts and screw fasteners. Even then, the moment that is induced on the shoe is so high during the test that it causes a rocking action. The degree of the rocking action depends on the temperature at which the test is carried out, thus indicating that the cause of the rocking is a machine compliance issue. Furthermore, many connections and bearings are located between the actuator and the MAST, which also adds deformation to the actuator linear variable differential transformer (LVDT) measurements. At a glance, an on-specimen LVDT should be a quick solution, but the rocking action makes those measurements inaccurate. Hence, the on-specimen displacement of the shear test is measured using an external non-contact DIC system.
5
4
3
2
MTS Actuator Displacement Displacement (mm) 1 DIC
0 0 10 20 30 40 50 60 Time (s)
Figure 3.36 Comparison of MTS actuator and DIC system displacements. The accuracy of a DIC system relies on the system set-up, which includes lighting, speckle pattern, aperture opening, etc. (Safavizadeh et al. 2017). Therefore, the DIC system’s accuracy must be checked carefully before conducting shear tests. This aim can be achieved by conducting a trial test at a constant displacement rate without any attached load using the DIC system and then comparing the results to the recorded actuator displacement(s). The entire measurement system settings should mimic real shear test conditions, except that the speckled paper that is used for tracking motion is glued to the MTS actuator. The environmental chamber also is used in this validation test because, during a monotonic shear test, the image of the test
53 specimen is captured through the transparent window of the environment chamber. The refraction of light through the window influences the DIC measurements. During the test, a constant displacement rate of 5.08 mm/min (0.2 in./min) is applied by the MTS machine. Images are taken by the DIC camera at a constant capture rate of 150 milliseconds. Figure 3.36 shows the displacements that were measured using an actuator and DIC system in this study. As shown, the DIC system measurements match the actuator displacements, thus validating the accuracy of the DIC system for taking displacement measurements during MAST tests.
54 4. PAVEMENT RESPONSE ANALYSIS
The presence of sufficient interface bond strength among the pavement layers is evaluated by understanding and quantifying the distribution of the stresses within the pavement section under realistic traffic conditions. NCSU research group has developed a fast Fourier transform-based three-dimensional (3D) viscoelastic finite element (FE) analysis tool known as FlexPAVE™ (formerly known as the LVECD program) to evaluate the pavement response under moving vehicle loads. It can simulate actual climatic conditions as generated by the Enhanced Integrated Climatic Model (EICM). Besides, the software is zipped with the Simplified ViscoElastic Continuum Damage (S-VECD) and permanent deformation shift model that could predict the pavement response and distresses viz. fatigue cracking and rutting for any temperature and any traffic conditions. In this study, FlexPAVE™ was used to determine the critical stresses involved in debonding. Cho (2017a) has conducted extensive pavement response analysis on three typical pavement sections constructed in North Carolina, categorized as thin, intermediate, and thick structures. The analysis was carried out at 5C, 20C, 40C, and 60C, three different speeds, 8 km/hour (5 mph), 40 km/hour (25 mph), and 88 km/hour (55 mph), three axle loads, 106.8 kN (24 kips), 160 kN (36 kips), and 213.6 kN (48 kips), and two types of tire rolling conditions, i.e., free-rolling and braking to determine the critical debonding condition. The outcome shows that the most critical stress state condition that leads to debonding is created by a thick pavement structure ran over by a single tire with a single – axle single - tire load of 213.6 kN (48 kips) at a fixed vehicular speed of 8 km/hour (5 mph) under braking condition. Henceforth, that specific condition is considered for the current study except that the tire loading is assumed as 80 kN (18 kips) single-axle dual-tire configuration.
4.1 Parameters Used in FlexPAVE™ Simulation 4.1.1 Structure Information Among the typical pavement sections in North Carolina, a thick pavement has higher vulnerable chances to debonding (Cho 2016). Figure 4.1 shows the cross-section view and provides the thickness of each layer assumed for the thick pavement structure. The top three layers i.e., surface, intermediate, and base layers, are asphalt concrete mixtures with different gradations. The overlay structure is considered in this study. Figure 4.2 presents the three
55 structures used in this analysis. The standard thickness for surface course constructed with asphalt mixture is having 9.5 mm nominal maximum aggregate size usually ranges between 38.1 mm (1.5 in.) and 63.5 mm (2.5 in.). Henceforth, in the present study, a thickness of 38.1 mm (1.5 in.) was chosen to analyze the critical condition.
Z
RS9.5C X 1.5 in 1.5
Y
I19B 2.5 in 2.5
B25B 6.0 in 6.0
Subgrade
Figure 4.1 Cross-section of thick pavement structure (X-axis: transverse direction and Y-axis: longitudinal/traffic direction)
Overlay RS9.5C
RS9.5C 1.5 in 1.5 Overlay RS9.5C
I19B Overlay RS9.5C 2.5 in 2.5 RS9.5C
1.5 in 1.5 RS9.5C 1.5 in 1.5 I19B I19B in 2.5
2.5 in 2.5 RS9.5C 2.5 in 2.5
ABC in
B25B in 8.0 6.0 in 6.0
.0 .0 ABC 6
Subgrade Subgrade Subgrade
(a) (b) (c) Figure 4.2 Pavement structure input: (a) thick structure, (b) intermediate structure, and (c) thin structure.
56 4.1.2 Material Parameters for Each Pavement Layers Asphalt Concrete The three top layers of the thick pavement structure are assigned with the material properties of asphalt concrete. The surface course is assigned with the RAP-40 mixture while the intermediate and base layers with properties of mixtures I19B and B25B, respectively. Table 4.1 Prony coefficients for relaxation modulus.
i Ei (MPa) (sec) S9.5C I19B B25B 2.00E+08 46.25 15.86 10.71 2.00E+07 28.63 31.18 21.24 2.00E+06 83.97 65.98 45.3 2.00E+05 202.61 150.13 105.76 2.00E+04 533.62 356.33 268.88 2.00E+03 1414.45 823.24 702.8 2.00E+02 3534.75 1685.67 1676.39 2.00E+01 7775.77 2821.99 3194.19 2.00E+00 14342.18 3759.34 4550.7 2.00E-01 21764.79 3476.25 4160.3 2.00E-02 27406.48 3756.6 4357.92 2.00E-03 29391.90 3081.01 3337.35 2.00E-04 27736.27 2440.15 2452.81 2.00E-05 23763.77 1802 1669.18 2.00E-06 18977.34 1286.44 1099.82 2.00E-07 14398.53 893.45 705.84 2.00E-08 10955.97 610.14 446.61
E∞ 14.04 38.24 51.59
The viscoelastic nature of asphalt concrete is defined with the aid of prony series coefficients/parameters in the FlexPAVE™. It is achieved through the interconversion from dynamic modulus to relaxation modulus over a wide time region using the generalized Maxwell model using the Eq. (4). The number of Maxwell elements decides the prediction accuracy, more elements, high accuracy but leads to more complexity at the same time. FlexMATTM, an excel based software, is used to analyze the dynamic modulus outcomes to calculate the Prony series coefficients, as shown in Table 4.1. The material properties for I19B and B25B were adopted from Cho (2016).
57
m −t i EtEEe()=+ i (4) i=1 where E(t) = the relaxation modulus,
E = the equilibrium modulus,
Ei = the relaxation strength,
i = the relaxation times, and m = the number of Maxwell elements. Subgrade The subgrade is assumed as linear elastic material in the FlexPAVE™ simulations and the modulus value used in the current analysis is 68.95 MPa (10,000 psi). 4.1.3 Climate Data Although FlexPAVE™ has the ability to simulate the pavement behavior under changing temperature as a function of time and pavement depth, the isothermal temperature profile at 50C was used in this study as it acts as the critical condition for debonding. 4.1.4 Traffic Data The design vehicle configuration for the response analysis is chosen as a dual-tire system to replicate the tire loading condition of a half of a single-axle dual-tire condition. An axle load of 80 kN is used, which is distributed through the dual tire configuration as 40 kN (9 kips) with 827.4 kPa (120 psi) tire-pavement contact pressure. 4.1.5 Tire-Pavement Contact Pressure Configuration The tire-pavement contact pressure distribution is non-uniform and mimicking it is essential in accurate pavement response computations. Moreover, the tire-pavement contact pressure distribution is affected significantly by tire inflation pressure, tire type, and tire load. NCSU research team determined the FlexPAVE™ tire-pavement contact area on the Stress-In- Motion (SIM) technology under the moving load (De Beer et al. 2004). The rectangular shape with an aspect ratio of 11/7 (length/width) is assumed in FlexPAVE™. The tire-pavement contact pressure distribution is based on fitting a quadratic function to the actual pressure in both the longitudinal and transverse directions. Figure 4.3 shows the FlexPAVE™ dual tire-pavement contact configuration.
58
Traffic Direction 21.44 cm (8.44 in) FlexPAVE™ Tire Configuration
9.04 cm (3.56 in) Axle Load: 40 kN Dual Tire
Tire Pressure: 827.37 kN 9.75 cm (7.67 in)
Transverse
Direction 19.49 cm (7.67 in) (7.67 cm 19.49
15.24 cm (6 in)
30.47 cm (12 in) 12.40 cm (4.88 in)
Figure 4.3 FlexPAVE™ dual tires-pavement contact configuration. 4.1.6 Shear Traction Due to Tire Braking Condition The tire-pavement contact stress is affected directly by the tire rolling conditions and thereby the stress response. Cho (2017a) considered two types of tire rolling conditions .i.e., free-rolling and braking. It is found from the study conducted by the National Highway Traffic Safety Administration (NHTSA) on the stopping distances of truck tractors that the rolling resistance coefficient (shear traction) varies from 0.35 to 0.55. The pavement response analysis shows that 0.55 shear traction as the most critical condition for debonding and hence used for the current analysis.
4.2 FlexPAVE™ Analysis Output 4.2.1 Stress and Strain Distribution FlexPAVE™ analysis of interface stress distribution at a depth of 3.81 cm (1.5 in.) for the thick structure is plotted in Figure 4.4. The three-dimensional (3-D) contour plots help to identify the critical conditions at the interface. OriginPro 2018 was employed to represent 3D contours and 2D contours. The contour profile option in the software helps to draw the response of stress/strain along transverse/longitudinal direction, as shown in Figure 4.5. The maximum normal stress (zz-max) is 740 kPa. Meantime, the magnitude of resultant shear stress at any location along the interface plane is calculated using Equation (8), and the maximum resultant shear stress (max) of 303 kPa is found to occur at the center of the tire.
59
800
Normal Stress (kPa) 600
400
200 20 Distance-30 from Center of Tire 10 in Transverse-20 Direction (cm) -10 0 0 10 -10 20 -20 30 Distance from Center of Tire in Longitudinal Direction (cm)
(a)
400
Shear Stress (kPa) 300
200
100 20 Distance-30 from Center of Tire 10 in Transverse-20 Direction (cm) -10 0 0 10 -10 20 -20 30 Distance from Center of Tire in Longitudinal Direction (cm)
(b) Figure 4.4 Stress distribution at the interface 1.5-inch deep: (a) normal stress and (b) shear stress.
60
800 600 400
200 Normal Stress (kPa)
200
400
600
800
Normal Stress (kPa) 0 -30 -20 -10 0 10 20 30 0 20 700
600 10 500
400 0 0 300
200 -10
100 Distance from Center of Tire in Longitudinal Direction (cm)
0 -20 15.24 Distance from Center of Tire in Transverse Direction (cm)
(a) 400 300 200
100 Shear Stress (kPa)
100 200 300 400
ShearStress (kPa) 0 -30 -20 -10 0 10 20 30 0 20
300 10 250 8.71
200 0 150
100 -10
50 Distancefrom CenterTire of in in Longitudinal Direction (cm)
0 -20 15.24 Distance from Center of Tire in Transverse Direction (cm)
(b) Figure 4.5 (a) Normal stress and (b) resultant shear stress distribution along the critical sections. The shear strains occurred along the longitudinal and transverse direction were computed and are shown in Figure 4.6 (a) and (b), respectively.
61
0.008
Shear Strain-γ 0.006
0.004
yz
(γ) 0.002 20 Distance-30 from Center of Tire 10 in Transverse-20 Direction (cm) -10 0 0 10 -10 20 -20 30 Distance from Center of Tire in Longitudinal Direction (cm)
(a)
0.004
Shear Strain-γ 0.002
0.000
zx
(γ) -0.002 20 Distance-30 from Center of Tire 10 in Transverse-20 Direction (cm) -10 0 0 10 -10 20 -20 30 Distance from Center of Tire in Longitudinal Direction (cm)
(b) Figure 4.6 Shear strain (a) yz and (b) zx.
62 5. INTERFACE SHEAR STRENGTH
A typical MAST test procedure allows a specimen to condition for three hours at a specific temperature prior to the commencement of the test. The geosynthetic-reinforced specimens were tested at three different temperatures, 23C (73F), 35C (95F), and 54C (129F), at a constant crosshead displacement rate of 5.08 mm/min (0.2 in./min). Previous research experience shows that conducting three replicate tests at 35C (the intermediate temperature) increases the accuracy of the fitted mastercurve. Only one specimen was tested at the remaining two temperatures, 23C and 54C. The actuator displacement rate was selected as 5.08 mm/min (0.2 in./min), which allows the monotonic shear test of the geosynthetic-reinforced specimen to complete quickly and to reach peak stress (shear strength) within 10 seconds. The highest debonding potential is the location of the maximum shear ratio (MSR) beneath a wheel load. The normal confining pressure experienced at the MSR point was found to be 172.37 kPa (25 psi) and was selected as the normal stress for most of the MAST tests. This finding was based on Cho’s no tack coat equation (Cho 2016). In addition, 275.79 kPa (40 psi) and 482.63 kPa (70 psi) were selected for the current study to see the effects of confinement on ISS outcomes. During the test, the normal stress was maintained using a confining plate. The confinement plate was tightened against the specimen until the required normal stress level was achieved prior to the start of the test. Throughout the test, the shear force and normal forces were continuously recorded. In addition, the DIC technique was used to measure the on-specimen displacements.
5.1 Shear Strain Rate Chehab et al. (2002) proposed DIC strain rate calculation method was used in this study. As shown in Figure 5.1, the DIC shear strain rate was determined by the linear potion, which follows the pure power law of the DIC strain.
63
Figure 5.1 DIC strain rate calculation method (Chehab et al. 2002). 5.2 Interface Shear Strength 5.2.1 Effects of Geosynthetic Interlayer Type on Interfacial Shear Strength The mastercurves were constructed for different geosynthetic-reinforced specimens based on tests carried out at three different temperatures and a constant actuator displacement rate of 5.08 mm/min (0.2 in./min). Figure 5.2 shows that the presence of any type of geosynthetic products at any testing conditions reduces the ISS and increases the chances of debonding. It is clear to distinguish the MAST specimens in this study into three categories based on the shear strength mastercurves. The control specimens show the best shear resistance in comparison to any geosynthetic reinforced specimens at any specific testing conditions. PC#1 and PaG display the higher shear strength among the geosynthetic-reinforced specimens, while PM, PF, and PC#2 show lower shear strength in comparison to the former category. Table 3.2 in Chapter 3 presents tack coat application rates applied for different geosynthetic interlayer products. The optimum tack coat application rates for PC#2 and PF are 1.36 L/m2 (0.3 gal/yd2) and 1.04 L/m2 (0.23 gal/yd2), respectively. PC#2 and PF exhibit greater asphalt retention capacity as they are the thickest products among the five, thereby demanding a high tack coat application rate. However, the effective tack coat application rate, which is the total tack coat application rate minus the asphalt retention rate, remains the same for all geosynthetic materials for each of the dry, optimum, and wet conditions. Thus, the major cause for the weak ISS of PC#2 and PF is the greater geosynthetic thickness of 2-3 mm (0.08-0.12 in.) compared to the average thickness of 1 mm (0.04 in.) for the other products. Consequently, these
64 factors force the reinforced interface to act as a softer shearing plane, which aids an easy failure. The viscoelastic nature of asphalt makes the tack coat-impregnated geosynthetic interface to behave as a stiffer interface layer at 23C (73F) than at the higher temperatures of 35C (95F) and 54C (129F). This viscoelasticity causes geosynthetic products such as PC#1 and PaG with similar thicknesses to yield comparable ISS values at 23C (73F). However, Figure 5.2 shows that the PaG reinforced specimen has the highest shear strength values among all the geosynthetic-reinforced specimens. The grid openings in the PaG product allow direct contact between the upper and lower asphalt concrete layers, which activates an additional interlocking action due to friction. The other geosynthetic products are continuous structures that avoid direct asphalt concrete layer contact. These products demand a high asphalt application rate compared to PaG’s tack coat application rate of 0.36 L/m2 (0.08 gal/yd2); thus, those products behave as a softer interlayer at higher temperatures. The temperature effect on the tack coat has less impact on the PaG reinforced specimens compared to the other specimens. The combination of these two factors (grid opening size and temperature dependency) explains the high ISS value for the PaG reinforced specimens at high temperatures.
2000 TC-Opt-0.03 y = 1448.3x0.10 y = 1582.4x0.17 PC#1-Opt-0.14 R² = 0.99 R² = 0.99 1600 PC#2-Opt-0.33 PF-Opt-0.23 y = 1583.6x0.18 y = 640.04x0.08 PaG-Opt-0.08 R² = 0.91 R² = 0.87 PM-Opt-0.13 1200 y = 1158.7x0.18 R² = 0.90 800
Shear Strength (kPa) Strength Shear 400 y = 889.31x0.15 R² = 0.91 0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
Figure 5.2 Mastercurve for different geosynthetic-reinforced asphalt specimens.
65 5.2.2 Effects of Tack Coat Application Rate on Interface Shear Strength Geosynthetic-reinforced MAST specimens with three different application rates that correspond to dry, optimum, and wet conditions were tested to measure the ISS. As shown in Figure 5.3, there do not have a clear indication to determine the tack coat application rate effect on geosynthetic material. The effect of the tack coat rates on the ISS, based solely on visual observation of the graphical representation of the results, is difficult to determine. Further statistical analysis could be useful in assessing the effect of the application rate.
66
Figure 5.3 Tack coat application rate effect on geosynthetic-reinforced specimen: (a) PC#1, (b) PaG, (c) PC#2, (d) PF, and (e) PM.
67
2000 PC#1_Wet_0.16 PC#1_Opt_0.14 1600 PC#1_Dry_0.12 y = 1687.5x0.19 R² = 0.99 1200 y = 1582.4x0.17 R² = 0.99 800
y = 1123.6x0.11 Shear Strength (kPa)ShearStrength 400 R² = 0.99
0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
(a) 2000 PaG_Wet_0.10 PaG_Opt_0.08 1600 PaG_Dry_0.06
1200 y = 1643.5x0.16 R² = 0.96 800 y = 1583.6x0.18 R² = 0.91 Shear Strength (kPa) Strength Shear 400 y = 1890.2x0.21 R² = 0.89 0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
(b)
68
2000 PC#2_Wet_0.35 PC#2_Opt_0.33 1600 PC#2_Dry_0.31
1200 y = 889.31x0.15 y = 597.56x0.09 R² = 0.91 R² = 0.99 800
Shear Strength (kPa)ShearStrength 400 y = 619.44x0.09 R² = 0.94 0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
(c) 2000 PF_Wet_0.25 PF_Opt_0.23 1600 PF_Dry_0.21
1200 y = 915.95x0.09 y = 640.04x0.08 R² = 0.99 R² = 0.87 800
400 Shear Strength (kPa) Strength Shear y = 1044.3x0.10 R² = 0.83 0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
(d)
69
2000 PM_Wet_0.15 PM_Opt_0.13 1600 PM_Dry_0.11
y = 1229x0.11 1200 R² = 0.75 y = 1078.1x0.11 R² = 0.97 800
0.18
Shear Strength (kPa)ShearStrength 400 y = 1158.7x R² = 0.90
0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
(e)
70 5.2.3 Effects of Temperature on Interface Shear Strength The MAST tests were conducted using five different geosynthetic products at three different temperatures to find the effects of temperature on the ISS. Figure 5.4 presents the comparisons. At the intermediate temperature (35C), at least two shear replicate tests were conducted to check for test variability. For the same type of geosynthetic-reinforced specimens, the shear strength decreases with an increase in temperature. Note that the effect of geosynthetic type on the ISS is evident at the low temperature (23C) but is nullified at the high temperature (54C). 5.2.4 Effects of Confining Pressure on Interface Shear Strength Three different normal (confining) stress, 172.37 kPa (25 psi), 275.79 kPa (40 psi), and 482.63 kPa (70 psi) were selected to conduct MAST test. The normal stress was applied by tightening the confinement plate against the specimen prior to the commencement of the test. The change in confinement load during the test was recorded using a load cell, The effect of confining pressure on the geosynthetic-reinforced specimens is plotted in Figure 5.4. It is apparent from the outcomes that an increase from 172.37 kPa to 482.63 kPa confining pressure increases the shear strength irrespective of the geosynthetic type used for specimen fabrication. However, PC#1 displays a comparable shear strength at a confining pressure of 172.37 kPa and 275.79 kPa while PM shows similar shear strength at a confining pressure of 275.79 kPa and 482.63 kPa. Figure 5.5 shows the correlation between shear strength and confining pressure found in this study. The PC#1 and PM results show that the effect of confinement pressure has a non-linear trend on ISS. This trend is against the typical linear trend found by previous researchers. Canestrari’s research group extensively studied the relationship between confining pressure and shear strength of geosynthetic-reinforced specimens (Canestrari et al. 2016, 2018, Ferrotti et al. 2011, Pasquini et al. 2013). They found that the ISS envelops with confining pressure follow a linear trend. Also, Cho and Kim (2015) found similar linear trend among shear strength and confining pressure at different strain rates during the ISS study on specimens fabricated using various tack coat emulsions. The limited data set available in the current study for confining pressure causes difficulty in ascertain the credibility of the non-linear trend. Non- typical trend could be attributed to the test variability. Therefore, considering the experience of previous researchers, the linear fit is used to develop the ISS prediction model in this study. However, it is recommended to carry out ISS study with a wider spectrum of confining pressure
71 as a future research goal to establish a clear trend on the effect of confining pressure on ISS of geosynthetic reinforced specimens.
2000 PC#1_25 psi_0.14 PC#1_40 psi_0.14 y = 1929.9x0.15 1600 PC#1_70 psi_0.14 R² = 0.93
0.14 1200 y = 1332.2x R² = 0.96
800 y = 1582.4x0.17 R² = 0.99 Shear Strength (kPa)ShearStrength 400
0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
(a) 2000 PM_25 psi_0.13 PM_40 psi_0.13 1600 PM_70 psi_0.13 y = 1335.3x0.12 R² = 0.99 1200 y = 1384.9x0.13 R² = 0.80 800 y = 1158.7x0.18
ShearStrength (kPa) 400 R² = 0.90
0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
(b) Figure 5.4 Confining pressure effect on geosynthetic-reinforced specimens: (a) PC#1 and (b) PM.
72
1600 23°C
1200 35°C 54°C
800
ShearStrength(kPa) 400
0 0 100 200 300 400 500 600 Confining Pressure (kPa)
(a) 1600 23°C
1200 35°C 54°C
800
Shear Strength (kPa) Strength Shear 400
0 0 100 200 300 400 500 600 Confining Pressure (kPa)
(b) Figure 5.5 Correlation between shear strength and confining pressure: (a) PC#1 and (b) PM (The numbers in the legend indicate the reduced strain rate).
73 5.3 Statistical Analysis on the Effect Tack Coat Application Rate The influence of the tack coat application rate on the bond strength was statistically analyzed using analysis of covariance (ANCOVA) model. Statistical software JMP 14 was used to conduct the single-factor ANCOVA model, as shown in Equation (5).
Yxxi ji ji=+−+ j (..) (5) where 휇 = the overall mean, i = 1, 2, 3, …, r, j = 1, 2, 3, …, n, β = regression coefficient for the relationship between the response and covariate,
xij = covariates, and
ϵij = the random errors. The result of the ANCOVA is present in Table 5.1. In this analysis, the tack coat application rate is the response, reduced strain rate is listed as the covariate. At α = 0.05, PM and PF found the tack coat application rate has a significant difference. For the other geosynthetic types, the tack coat application rate does not show statistical significance. Further, Tukey HSD analysis conducted on PM and PF, as shown in Table 5.2 and Table 5.3, respectively. The p- values shown in these tables indicate that the rate changes from dry to optimum has significant effects on the ISS of PM and PF. Also, the effect of rate changing from optimum to wet on ISS is significant for PF, but not for PM. Comparison of ISS measured at wet and optimum application rates as well as the ISS at dry and wet application rates shows an insignificant effect (p>0.05) of application rates on ISS for PM. Also, the rate change from dry to wet shows an insignificant effect on ISS for PF. Table 5.1 ANCOVA analysis summary. Geosynthetic Type Sum of Squares F Ratio Prob > F PC#1 0.001 2.31 0.19 PC#2 0.001 0.08 0.92 PM 0.076 6.30 0.03 PF 0.065 19.57 0.001 PaG 0.007 0.58 0.58
74 Table 5.2 Tukey HSD analysis for PM. Level - Level Difference Std Err Dif Lower CL Upper CL p-Value Dry Opt 0.196 0.057 0.028 0.364 0.0259 Wet Opt 0.122 0.057 -0.046 0.289 0.1505 Dry Wet 0.074 0.063 -0.112 0.261 0.5044
Table 5.3 Tukey HSD analysis for PF. Level - Level Difference Std Err Dif Lower CL Upper CL p-Value Dry Opt 0.169 0.030 0.081 0.257 0.0019 Wet Opt 0.136 0.030 0.049 0.224 0.0063 Dry Wet 0.033 0.033 -0.065 0.131 0.6065
Figure 5.6 shows a pictorial representation of the effect of tack coat application rate on ISS for PM and PF. The data shown in Figure 5.6 do not follow the expected trend, i.e., the optimum tack coat application rate is expected to yield greater shear strength than the dry and wet rates do. Within the data generated in this study, it is difficult to find reasons for this unexpected trend. Further study involving more tests is needed to explain this unexpected behavior.
75
1000
800
600
400
Shear Strength (kPa)ShearStrength 200
0 Wet Opt Dry Tack Coat Application Rate (a) 1000
800
600
400
ShearStrength (kPa) 200
0 Wet Opt Dry Tack Coat Application Rate
(b) Figure 5.6 Pictorial representation of the effect of tack coat application rate on ISS: (a) PM, (b) PF. 5.4 Shear Strength Prediction Model Cho (2016) utilized Modified Asphalt Shear Tester (MAST) with various temperatures, loading rates, and normal confining stresses to construct a shear strength prediction model, as shown in Equation (6). This model equation is embedded with the time-temperature superposition principle, which could predict the shear strength of asphalt concrete pavements at any strain rate and temperature combinations as well as at any normal confining stresses. Table
76 5.4 shows the coefficients for PC#1 and PM shear strength prediction model at the reference temperature of 35C. In this study, this prediction model is applied to conduct preliminary analysis. Figure 5.7 presents the validation results of the PC#1 prediction model at 172 kPa confining pressure using wet and dry tack coat application conditions. The R2 calculated from the validation is 0.803.
db f=c && R +(( a R ) + e ) c (6) where f = shear strength at the layer interface, kPa, & R = reduced shear strain rate, and c = normal confining stress, kPa. Table 5.4 Coefficients for PC#1 and PM shear strength prediction models. Layer Interface Condition a b c d e R2 PC#1 227.80 2.48 893.70 0.13 0.55 0.99 PM 44.92 14.17 1130.00 0.22 0.77 0.82
1200
1000
800
600
400
200 Perdicted Shear Strength (kPa) Strength PerdictedShear 0 0 200 400 600 800 1000 1200 Measured Shear Strength (kPa)
Figure 5.7 PC#1 prediction model validation at 172 kPa confining pressure.
5.5 Shear Ratio and Bonding Failure Definition
Shear ratio (SR) was calculated by the ratio of the computed shear stress (τmax) to the shear bond strength (τs), as shown in Equation (7), which could be used as an indicator to
77 determine interface bonding integrity. The shear stress (τmax) distribution was computed from the FlexPAVE™ analysis program, as shown in Equation (8). Meanwhile, shear bond strength was obtained from the shear strength prediction model developed based on the MAST outcomes. The input parameters for normal stress and shear strain rate for the strength prediction model are corresponding to a specific location of computed shear stress. Shear Ratio (SR)= (7) s where
τmax = FlexPAVE™ computed shear stress, kPa,
τs = laboratory test-based shear strength prediction model, kPa.
22 =+ zxyz (8) where = magnitude of resultant shear stress, zx = FlexPAVE™ computed shear stress in the transverse direction, kPa, and yz = FlexPAVE™ computed shear stress in the longitudinal direction, kPa. Figure 5.8 shows the shear strain on the longitudinal direction (yz) and the transverse direction (zx). The region within the yellow dotted line means the tire imprint is on the evaluation point. The lower boundary shows the front edge of the moving tire touches the evaluation point in the FlexPAVE™. The shear strain rate was calculated from the slope of the shear strain. Transverse and longitudinal shear strain were obtained separately and their resultant was further input into the shear strength prediction equation.
78 Time
Evaluation Point in FlexPAVE 0.008 50_1_Eyz Center 0.006
0.004 Shear Strain Shear 0.002
0 0.00 1.00 2.00 3.00 4.00 5.00 6.00 Simulation Time (s)
(a) 0.008 50_1_Ezx 0.006
0.004
Shear Strain Strain Shear 0.002
0.000
-0.002 0.00 1.00 2.00 3.00 4.00 5.00 6.00 Simulation Time (s)
(b) Figure 5.8 Shear strain output: (a) Eyz and (b) Exz
79 Maximum Shear Ratio (MSR) was used as the shear failure criterion in this study. Higher the MSR value is, more likely shear failure would occur at the pavement interface. Figure 5.9 provides guidance for the selection of the appropriate level of normal confining stress on the central longitudinal axis of the tire at the layer interface that corresponds to the location of the MSR.
1000 1.0 Normal Stress Shear Stress 800 Strength 0.8
Shear Ratio Shear Ratio Shear 600 0.6
400 0.4 Stress(kPa) 200 0.2
0 0.0 -0.3 -0.1 0.1 0.3 Longitudinal Distance from the Center of the Tire (m)
Figure 5.9 PM Shear ratio, shear strength, and shear and normal stress levels in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN (9 kips), 1.6 km/hour (1 mph), 50C, at 3.81 cm (1.5 in.) depth, and braking condition.
80
1.0 1.0
0.8 0.8
Shear Ratio Shear Ratio 0.6 0.6
0.4 0.4
0.2 0.2 20 20
Distance-30 from Center of Tire 10 Distance-30 from Center of Tire 10 in Longitudinal-20 Direction (cm) in Longitudinal-20 Direction (cm) -10 0 -10 0 0 0 10 -10 Direction (cm) 10 -10 Direction (cm) 20 -20 20 -20 30 30 DistanceTransverse from Center of Tire DistanceTransverse from Center of Tire in in
(a) (b) 0.8 0.8 0.6 0.6
0.4 0.4 Shear Ratio Shear 0.2 Ratio Shear 0.2 Shear Ratio Shear Ratio
0.0 0.0
0.2 0.2
0.4 0.4
0.6 0.6 0.0 0.8 0.0 0.8 -30 -20 -10 0 10 20 30 -30 -20 -10 0 10 20 30 20 20 0.6
0.5 0.6 11.18 11.64 10 10 0.5 0.4 0.4 0.3 0 0 0.3 0.2 0.2
-10 -10
Distance from Center of Tire of Center from Distance Distance from Center of Tire of Center from Distance
0.1 (cm) Direction Longitudinal in 0.1 (cm) Direction Longitudinal in
0.0 -20 0.0 -20 15.24 15.24 Distance from Center of Tire Distance from Center of Tire in Transverse Direction (cm) in Transverse Direction (cm) (c) (d) Figure 5.10 Shear ratio distribution under dual tires in 3D space (a) PC#1 condition, (b) PM condition and in 2D space contour (c) in PC#1 condition, and (d) PM condition. Figure 5.10 shows the shear ratio distribution under dual tire at the interface. For the specific condition considers in this study, MSR occurs at a point in front of the tire along the center-line of the tire, as shown in Figure 5.10 (c) and (d). It is also evident that the PM shows a higher MSR than PC#1 reinforced pavements. This indicates that PM reinforced pavement has a higher potential to occur debonding failure. Figure 5.11 presents the pavement structure effect on the shear ratio. It is found that a similar shear ratio trend among thin, intermediate, and thick structures. The comparison of MSR among three different structures shows that the thick structure has a slightly higher MSR value (Figure 5.12). However, the difference in the MSR values is so small (2.5% - 3.5%) among different structures that the pavement structures considered for the current study do not have a significant effect on MSR. This observation is because shear debonding is a near-surface phenomenon. The temperature effect on the shear ratio
81 is shown in Figure 5.13. The shear ratio increases with the rise of the temperature, Figure 5.14 shows the highest MSR value happens at the highest simulation temperature, 50C. This finding is due to the decrease of the ISS at higher temperatures, while the shear stress at the interlayer remains constant (fully bonding condition is assumed) irrespective of the temperature. Figure 5.15 compares the effect of overlay thickness on the shear ratio. It is apparent that the thinnest overlay (3.81 cm or 1.5 in) has a higher shear ratio along the interface in comparison to the other thicknesses. The comparison of MSR for different overlay thicknesses shown in Figure 5.16 reveals that one centimeter increases in the thickness reduces the MSR by 0.04. Note that the AC overlay thickness thinner than 3.81 cm (1.5 in.) is not typical and, therefore, not evaluated in this study. The effect of vehicular speed on the shear ratio was evaluated in Figure 5.17. It is a well- known fact that the loading rate influences AC strength due to its viscoelastic nature. The faster loading rate causes higher strength, while slower loading rate yields lower strength. Henceforth, Figure 5.18 displays that a vehicular speed of 1.6 km/h (1 mph) imparts the highest MSR. The slowest speed that could be simulated using FlexPAVE™ is 1.6 km/h (1 mph); however, a real field braking condition demonstrates in Figure 5.19 indicating that the lowest speed before the vehicle comes to a standstill is the most critical for the debonding at the layer interface. The comparison of all factors using PM and PC#1 products shows that PC#1 outperforms PM in all the analysis conditions. In summary, the highest debonding potential among the FlexPAVE™ simulation conditions in this study occurs in thick overlay structure, 50C, 1.6 km/h, and at 3.81 cm depth.
82
1 Thin 0.8 Intermediate Thick 0.6
0.4 ShearRatio
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(a) 1 Thin 0.8 Intermediate
0.6 Thick
0.4 Shear Ratio Shear
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(b) Figure 5.11 Effects of pavement structure on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: 40 kN, 1.6 km/hour, 50C, at 3.81 cm depth, and braking condition (a) PC#1 and (b) PM.
83
1.0 PM PC#1 0.8
0.6
MSR 0.4
0.2
0.0 Thick Intermediate Thin Structure
Figure 5.12 Effects of pavement structure on MSR: 40 kN, 1.6 km/hour, 50C, at 3.81 cm depth, and braking condition.
84
1 50°C 0.8 40°C 30°C 0.6
0.4 ShearRatio
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(a) 1 50°C 0.8 40°C 30°C 0.6
0.4 Shear Ratio Shear
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(b) Figure 5.13 Effects of temperature on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN, 1.6 km/hour, at 3.81 cm depth, and braking condition (a) PC#1 and (b) PM.
85
1.0 PM 0.8 PC#1
0.6 MSR 0.4
0.2
0.0 20 30 40 50 60 Temperature (C)
Figure 5.14 Effects of temperature on MSR: thick pavement, 40 kN, 1.6 km/hour, at 3.81 cm depth, and braking condition.
86
1 1.5 in 0.8 2 in 3 in 0.6
0.4 ShearRatio
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(a) 1 1.5 in 0.8 2 in 3 in 0.6
0.4 Shear Ratio Shear
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(b) Figure 5.15 Effects of overlay thickness on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN, 1.6 km/hour, 50C, and braking condition (a) PC#1 and (b) PM.
87
1.0 PM 0.8 PC#1
0.6
MSR 0.4
0.2
0.0 0 2 4 6 8 10 Overlay Thickness (cm)
Figure 5.16 Effects of overlay thickness on MSR: thick pavement, 40 kN, 1.6 km/hour, 50C, and braking condition.
88
1 1 mph 0.8 3 mph 5 mph 0.6 20 mph 45 mph
0.4 ShearRatio
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(a) 1 1 mph 0.8 3 mph 5 mph 20 mph 0.6 45 mph
0.4 Shear Ratio Shear
0.2
0 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 Longitudinal Distance from the Center of the Tire (m)
(b) Figure 5.17 Effects of vehicle speed on shear ratio in the longitudinal direction under the central axis of the tire at the layer interface: thick pavement, 40 kN, 50C, at 3.81 cm depth, and braking condition (a) PC#1; (b) PM.
89
1.0
0.8
0.6 MSR 0.4
0.2 PM PC#1 0.0 0 20 40 60 80 Speed (km/h)
Figure 5.18 Effects of vehicle speed on MSR: thick pavement, 40 kN, 50C, at 3.81 cm depth, and braking condition.
Brake Critical Condition
45 mph 40 mph 15 mph Near Stop Time
Figure 5.19 Critical condition schematic.
5.6 Acceptance Criterion for Debonding Resistant Geosynthetic Products The MSR critical conditions for PC#1 and PM are listed in Table 5.5. It is found that the critical MSR location lies slightly in front of the tire. In order to keep geosynthetic-reinforced pavement from debonding, its shear strength should be higher than the shear stress at critical condition, which means the MSR value should always less than 1. Figure 5.20 presents the temperature relationship with DIC reduced strain rate at the MTS actuator displacement rate of 5.08 mm/min (0.2 in./min), the corresponding temperature to MSR critical condition reduced
90 strain rate is around 50C. Therefore, the threshold shear test for evaluation of the geosynthetic products should be conducted at 50C, 5.08 mm/min deformation rate, and 275.8 kPa (40 psi) confining pressure. Based on the MSR information, the minimum required shear strength for the geosynthetic-reinforced specimen is 305 kPa (44 psi). It should be noted that the bonding at the interlayer will deteriorate with the environment influence and repeated traffic loading. Regarding the field condition, a safety factor should be taken into consideration. Due to the machine compliance, this threshold testing condition was determined by the relationship between on- specimen and loading frame displacement rates, thus only applicable to the MTS loading frame and MAST at NCSU. The laboratory intending to carry out this type of study should adjust the temperature based on the machine compliance to match the critical shear strain rate.
Table 5.5 MSR location information summary.
Normal Stress (kPa) 283.3 Shear Stress (kPa) 305.4 Shear Strain Rate 1.9E-04 PM 0.67 MSR PC#1 0.80 Distance to Center of the tire (cm) 10.56 Distance to tire front edge of the tire (cm) 0.81
100 y = 18.16x-0.12
80 R² = 1.00
C)
60
40 Temperature ( Temperature 20
0 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 DIC Reduced Strain Rate
Figure 5.20 Temperature relationship with DIC reduced strain rate at 5.08 mm/min MTS deformation rate.
91 6. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE STUDY
Based on the experimental investigations, the current research studied the effects of geosynthetic types, confining pressures, temperatures, and shear rates on geosynthetic-reinforced interlayer bonding performance and developed a shear strength prediction model and proposed geosynthetic acceptance criterion. The findings of this study are as follows: • The presence of any geosynthetic product at any testing condition reduces the ISS and increases the chances of debonding. The following three categories are developed based on their shear strengths. The control specimens (without an interlayer) show the best shear resistance in comparison to any geosynthetic- reinforced specimens at any testing conditions. Paving composite #1 (PC#1) and paving grid (PaG) display the higher shear strength among the geosynthetic- reinforced specimens, while paving mat (PM), paving fabric (PF), and paving composite #2 (PC#2) show lower shear strength in comparison to the former category. • The effect of the tack coat application rate on the ISS of geosynthetic-reinforced specimens is not clear. The statistical analysis of the ISS from the PM and PF reinforced specimens shows that the tack coat application rate changing from dry to optimum has significant effects on ISS of PM and PF. Also, the effect of rate changes from optimum to wet on ISS is significant for PF, but not for PM. The ISS values from PF and PM show an unexpected trend, i.e., ISS of the optimum rate is lower than ISS of dry or wet conditions. Within the data generated in this study, it was not possible to explain this unexpected trend. • For the same type of geosynthetic-reinforced specimen, the shear strength decreases with an increase in temperature. The effect of geosynthetic type on the ISS is evident at the low temperature (23C) but is nullified at the high temperature (54C). The shear strength is found to reduce 40-65% with a change in temperature from 23C to 54C. The shear strength variability is found to decrease with an increase in the testing temperature.
92 • The effect of confining pressure shows that an increase from 172.37 kPa to 482.63 kPa confining pressure increases the shear strength irrespective of the geosynthetic types. The FlexPAVE™ analyzed on various overlay pavement structures, speeds, temperatures, and overlay thicknesses suggest: • The difference in the MSR values among different structures structure is between 2.5% - 3.5%. Pavement structures considered for the current study do not have a significant effect on MSR as shear debonding is a near-surface phenomenon. The thick structure is selected as the representative structure. It is found that the higher temperature and lower speed condition yield highest debonding potential among the simulation conditions. The interlayer shear resistant ability is the weakest while the vehicle comes to a standstill. Also, the thinnest overlay (1.5 in) shows the highest MSR value. • The highest debonding potential among the FlexPAVE™ simulation conditions in this study occurs in the thick overlay structure, 50C, 1.6 km/h, and 3.81 cm depth. Also, according to MSR analysis the threshold shear strength test for evaluation of geosynthetic products should be conducted at 50C, 5.08 mm/min (0.2 in./min) deformation rate, and 275.8 kPa (40 psi) confining pressure. Based on the MSR information, the minimum required shear strength for geosynthetic-reinforced specimens at this condition is 305 kPa (44 psi). It should be noted that the bonding at the interlayer will deteriorate with the environment influence and traffic loading. Regarding the field condition, a safety factor should be taken into consideration. Due to the machine compliance, this threshold testing condition is only applicable to the MTS loading frame and MAST shear apparatus at NC State University. The laboratory intending to carry out this type of study should adjust the temperature based on the machine compliance to match the critical shear strain rate. Recommendations for future studies are as follows: • In this study, only preliminary tests were conducted to evaluate the tack coat application rate effect on ISS. However, the comprehensive tack coat application rate study on geosynthetic-reinforced materials should be carried out to determine the optimum tack coat application rates for each geosynthetic products.
93 • The monotonic shear test on geosynthetic-reinforced specimens could only be used as an index test to evaluate the debonding potential and a selection criterion for geosynthetic products. It is recommended to conduct the shear fatigue test to understand the service life of geosynthetic-reinforced interlayer. • The monotonic shear test could be used to evaluate the geosynthetic-reinforced field core specimens. Thus, the relationship between laboratory fabricated specimens and filed cores can be established.
94 REFERENCES
Al-Qadi, I. L., Carpenter, S. H., Leng, Z., Ozer, H., and Trepanier, J., 2008. Tack coat optimization for HMA overlays: Laboratory testing. Illinois Center for Transportation (ICT), Rantoul, Illinois, United States. Al-Qadi, I. L., Carpenter, S. H., Leng, Z., Ozer, H., and Trepanier, J., 2009. Tack coat optimization for HMA overlays: Accelerated pavement test report. Illinois Center for Transportation (ICT), Rantoul, Illinois, United States. American Association of State and Highway Transportation Officials (AASHTO)., 2012. Standard method of test for theoretical maximum specific gravity (Gmm) and density of hot mix asphalt (HMA). AASHTO, Washington, D.C., United States. American Association of State and Highway Transportation Officials (AASHTO)., 2017. Standard Practice for Preparation of Cylindrical Performance Test Specimens Using the Superpave Gyratory Compactor (SGC). AASHTO, Washington, D.C., United States. Amini, F., 2005. Potential applications of paving fabrics to reduce reflective cracking. Mississippi Department of Transportation, Jackson, Mississippi, United States. Bae, A., Mohammad, L. N., Elseifi, M. A., Button, J., and Patel, N., 2010. Effects of temperature on interface shear strength of emulsified tack coats and its relationship to rheological properties. Transportation research record, 2180(1), 102–109. Baek, J., 2010. Modeling reflective cracking development in hot-mix asphalt overlays and quantification of control techniques. PhD Dissertation, University of Illinois at Urbana- Champaign, Champaign, Illinois, United States. Blankenship, P., Iker, N., and Drbohlav, J., 2004. Interlayer and design considerations to retard reflective cracking. Transportation Research Record, 1896(1), 177–186. Bognacki, C. J., Frisvold, A., and Bennert, T., 2007. Investigation of Asphalt Pavement Slippage Failures on Runway 4R-22L, Newark International Airport. 2007 Worldwide Airport Technology Transfer Conference, Atlantic City, New Jersey, United States. Boulangé, L. and Sterczynskia, F., 2012. Study of interfacial interactions between bitumen and various aggregates used in road construction. Journal of adhesion science and technology, Taylor & Francis, 26(1–3), 163–173. Button, J. W. and Lytton, R. L., 1987. Evaluation of fabrics, fibers and grids in overlays. International Conference on the Structural Design, Ann Arbor, Michigan, United States, 925– 934. Button, J. W. and Lytton, R. L., 2007. Guidelines for using geosynthetics with hot-mix asphalt overlays to reduce reflective cracking. Transportation Research Record, 2004(1), 111–119. Canestrani, F., Grilli, A., Santagata, F., and Virgili, A., 2006. Interlayer shear effects of geosynthetic reinforcements. 10th International Conference on Asphalt Pavements, Quebec City, Canada. Canestrari, F., D’Andrea, A., Ferrotti, G., Graziani, A., Partl, M. N., Petit, C., Raab, C., and Sangiorgi, C., 2018. Testing and Characterization of Sustainable Innovative Bituminous
95 Materials and Systems. Advanced interface testing of grids in asphalt pavements, Springer International Publishing. Canestrari, F., Ferrotti, G., Abuaddous, M., and Pasquini, E., 2016. Geocomposite-reinforcement of polymer-modified asphalt systems. 8th RILEM International Symposium on Testing and Characterization of Sustainable and Innovative Bituminous Materials, Springer Netherlands, Dordrecht, Netherlands, 383–395. Chehab, G., Kim, Y., Schapery, R., Witczak, M., and Bonaquist, R., 2002. Time-temperature superposition principle for asphalt concrete with growing damage in tension state. Journal of the Association of Asphalt Paving Technologists, Association of Asphalt Paving Technologists (AAPT), Colorado Springs, Colorado, United States, 559–593. Chehab, G. R., O’Quinn, E., and Richard Kim, Y., 2000. Specimen geometry study for direct tension test based on mechanical tests and air void variation in asphalt concrete specimens compacted by superpave gyratory compactor. Transportation Research Record, 1723(1), 125– 132. Cho, S. H., 2016. Evaluation of interfacial stress distribution and bond strength between asphalt pavement layers. PhD Dissertation, North Carolina State University, Raleigh, North Carolina. Cho, S. H., Karshenas, A., Tayebali, A. A., Guddati, M. N., and Kim, Y. R., 2017a. A mechanistic approach to evaluate the potential of the debonding distress in asphalt pavements. International Journal of Pavement Engineering, 18(12), 1098–1110. Cho, S. H. and Kim, Y. R., 2016. Verification of Time–Temperature Superposition Principle for Shear Bond Failure of Interlayers in Asphalt Pavements. Transportation Research Record: Journal of the Transportation Research Board, 2590(1), 18–27. Cho, S. H., Safavizadeh, S. A., and Kim, Y. R., 2017b. Verification of the applicability of the time–temperature superposition principle to interface shear stiffness and strength of GlasGrid- reinforced asphalt mixtures. Road Materials and Pavement Design, 18(4), 766–784. Christensen, D. W. and Anderson, D. A., 1992. Interpretation of dynamic mechanical test data for paving grade asphalt cements (with discussion). Journal of the Association of Asphalt Paving Technologists, 61–116. Collop, A., Sutanto, M., Airey, G., and Elliott, R. C., 2011. Development of an automatic torque test to measure the shear bond strength between asphalt. Construction and Building Materials, Elsevier, 25(2), 623–629. De Beer, M., Fisher, C., and Kannemeyer, L., 2004. Towards the application of Stress-In-Motion (SIM) results in pavement design and infrastructure protection. 8th International Symposium on Heavy Vehicle Weights and Dimensions, Johannesburg, South Africa. Federal Highway Administration., 2016. Tack Coat Best Practices. FHWA, Washington, D.C., United States, 1–17. Ferrotti, G., Canestrari, F., Pasquini, E., and Virgili, A., 2012. Experimental evaluation of the influence of surface coating on fiberglass geogrid performance in asphalt pavements. Geotextiles and Geomembranes, Elsevier, 34, 11–18.
96 Ferrotti, G., Canestrari, F., Virgili, A., and Grilli, A., 2011. A strategic laboratory approach for the performance investigation of geogrids in flexible pavements. Construction and Building Materials, Elsevier Ltd, 25(5), 2343–2348. Hachiya, Y., Umeno, S., and Sato, K., 1997. Effect of tack coat on bonding characteristics at interface between asphalt concrete layers. Doboku Gakkai Ronbunshu, Japan Society of Civil Engineers, 1997(571), 199–209. Huesker., 2015. HaTelit® Asphalt Reinforcement Grid Installation Guideline.
97 Pasquini, E., Bocci, M., Ferrotti, G., and Canestrari, F., 2013. Laboratory characterisation and field validation of geogrid-reinforced asphalt pavements. Road Materials and Pavement Design, 14(1), 17–35. Pasquini, E., Giacomello, G., Pasetto, M., and Canestrari, F., 2015. Laboratory evaluation of the effect of low-temperature application of warm-mix asphalts on interface shear strength. Construction and Building Materials, Elsevier, 88, 56–63. Peattie, K. R., 1980. The incidence and investigation of slippage failures. The Performance of Rolled Asphalt Road Surfacings, Thomas Telford Publishing, 3–15. Raab, C. and Partl, M. N., 2004. Interlayer shear performance: experience with different pavement structures. 3rd Eurasphalt & Eurobitume Congress, Foundation Eurasphalt, Vienna, Austria, 535–545. Raab, C., Partl, M. N., and El Halim, O. A., 2009. Evaluation of Interlayer Shear Bond Devices for Asphalt Pavements. Baltic Journal of Road and Bridge Engineering, 4(4), 186–195. Raposeiras, A., Castro-Fresno, D., Vega-Zamanillo, A., and Rodriguez-Hernandez, J., 2013. Test methods and influential factors for analysis of bonding between bituminous pavement layers. Construction and Building Materials, Elsevier, 43, 372–381. Rigo, J., 1993. General introduction, Main conclusions of the 1989 conference on reflective cracking in pavements, and future prospects. Reflective cracking in pavements: state of the art and design recommendations, London: E & FN Spon, Liege, Belgium, 3–20. Safavizadeh, S. A. and Kim, Y. R., 2017. DIC technique to investigate crack propagation in grid- reinforced asphalt specimens. Journal of Materials in Civil Engineering, 29(6), 1–13. Sprague, C. J., Allen, S., and Tribbett, W., 1998. Tensile properties of asphalt overlay geosynthetic reinforcement. Transportation Research Record, 1611(1), 65–69. Su, K., Sun, L., Hachiya, Y., and Maekawa, R., 2008. Analysis of shear stress in asphalt pavements under actual measured tire-pavement contact pressure. Proceedings of the 6th ICPT, The Committee on Pavement Engineering, JSCE, Sapporo, Japan, 11–18. Sudarsanan, N., Karpurapu, R., and Amrithalingam, V., 2018. An investigation on the interface bond strength of geosynthetic-reinforced asphalt concrete using Leutner shear test. Construction and Building Materials, Elsevier, 186, 423–437. Sutanto, M. H., 2009. Assessment of bond between asphalt layers. PhD thesis, University of Nottingham, Nottingham, United Kingdom. Tashman, L., Nam, K., and Papagiannakis, A. T., 2006. Evaluation of the influence of tack coat construction factors on the bond strength between pavement layers. Washington State Department of Transportation, Pullman, Washington, United States. TenCate., 2019. Installation Guidelines for Tencate Mirafi® MPG Composite Paving Grids.
98 Tozzo, C., Fiore, N., and D’Andrea, A., 2014. Dynamic shear tests for the evaluation of the effect of the normal load on the interface fatigue resistance. Construction and Building Materials, 61, 200–205. Uzan, J., Livneh, M., and Eshed, Y., 1978. Investigation of adhesion properties between asphaltic-concrete layers. Association of Asphalt Paving Technologists Proc, Lake Buena Vista, Florida, 495–521. Vanelstraete, A. and De Bondt, A., 1997. Crack prevention and use of overlay systems. Prevention of Reflective Cracking in Pavements: State-of-the-art report of RILEM Technical Committee 157 PRC, Systems to Prevent Reflective Cracking in Pavements, E & FN Spon, London. Vismara, S., Molenaar, A., Crispino, M., and Poot, M., 2012. Toward a better understanding of benefits of geosynthetics embedded in asphalt pavements. Transportation Research Record, (2310), 72–80. Wilson, B., Arif Chowdhury, Sheng Hu, Kim, S.-S., Nazzal, M., and Abbas, A., 2017. Tack Coat Performance and Materials Study. Federal Highway Administration; Texas A&M Transportation Institute; Ohio Department of Transportation. Wilson, B., Seo, A. Y., and Sakhaeifar, M., 2016. Performance Evaluation and Specification of Trackless Tack. Texas Department of Transportation, Austin, Texas, United States. Zhiming, T., 1997. Mechanistic Analysis for Reflective Cracking in Asphalt Overlay. Journal of Tongji University Natural Science, 25(6), 635–639. Zhou, F. and Sun, L., 2000. Mechanistic analysis of reflective cracking and validation of field test. Fourth International RILEM Conference on Reflective Cracking in Pavements-Research in Practice, RILEM Publications SARL, 81–91. Zofka, A., Maliszewski, M., Bernier, A., Josen, R., Vaitkus, A., and Kleizienė, R., 2015. Advanced shear tester for evaluation of asphalt concrete under constant normal stiffness conditions. Road Materials and Pavement Design, 16(sup1), 187–210.