To evaluate the performance of Warm Mixture Asphalt (WMA) with varying amounts of recycled asphalt pavement (RAP) in comparison with Hot Mix Asphalt (HMA), comprehensive laboratory and field evaluations were conducted. Mix designs were performed for both WMA with a LEADCAP additive and HMA with large amounts of fractionated RAP materials. Hamburg Wheel Tracking (HWT) test was performed to evaluate the rutting and moisture susceptibility of both HMA and WMA laboratory mixtures. HMA mixtures with up to 50% RAP materials by binder replacement exhibited a better performance than WMA mixtures. However, when RAP materials were increased to 75% both WMA and HMA mixtures showed a superior performance. When a specially designed LEADCAP additive for a mixture with a high RAP content called "RAPCAP" was used, the performance was significantly improved. The existing Asphalt Bond Strength (ABS) test (AASHTO TP91-11) was modified to better evaluate the adhesion bond between asphalt binder and aggregate surface. Based on the modified ABS test results, it was found that the asphalt binder type significantly influenced the adhesion bond. To evaluate the performance of WMA mixtures in the field, test sections were constructed in Iowa, Minnesota and Ohio. The test sections were successfully constructed with less compaction effort than HMA and met the required field densities per each DOT's specification. All HMA and WMA mixtures collected from the test sections passed the HWT and the modified Lottman tests, which indicates high resistance to rutting and moisture damage. The asphalt binders were then extracted and recovered from the field samples then re-graded following AASHTO M320 and AASHTO MP19-10. The recovered asphalt binder grades were found to be higher than the target grades due to the existence of RAP materials in the mixtures except for asphalt binders extracted from WMA mixtures produced using "RAPCAP" additive.

The objective of this research was to investigate the structural capacity of aggregate base materials stabilized with triaxial geogrid placed in a full-scale pavement involving control, or unstabilized, sections. Field testing was performed on a roadway in northeastern Utah that was 16 km (10 miles) long and included 10 test sections, seven stabilized sections and three control sections, each having five test locations. The pavement structure was comprised of a hot mix asphalt layer overlying an untreated aggregate base layer of varying thickness, depending on the test section. Except for the control sections, one or two layers of geogrid were incorporated into portions of the pavement structure at different locations. Falling-weight deflectometer testing and dynamic cone penetrometer testing were used to evaluate the structural capacity of the aggregate base layer in each pavement section. For data analysis, the Rohde's method was applied in conjunction with the 1993 American Association of State Highway and Transportation Officials pavement design guide methodology, and the Area under the Pavement Profile (AUPP) method was applied in conjunction with a mechanistic-empirical pavement analysis. Statistical analyses were then performed to enable comparisons of the test sections. Field results indicated that the asphalt layer thickness was consistently 140 mm (5.5 in.) at all 10 test sections, and the base layer thickness varied from 360 mm (14 in.) to 510 mm (20 in.). The results of the statistical analyses indicated that the majority of the 45 possible pairwise comparisons among the test sections were not statistically significant, meaning that variations in the presence and position of triaxial geogrid at those sections did not appear to affect the structural capacity. The remaining comparisons, however, were statistically significant and involved the test sections with the highest structural capacity. While one of these was unexpectedly an unstabilized control section, the others were constructed using one or two layers of geogrid in the base layer. In addition to being statistically significant, the observed differences were also practically important. Increases in the observed base layer coefficient from 0.12 to 0.18 correspond to an increase in the allowable number of equivalent single axle loads (ESALs) from 5.9 million to 19.2 million at the research site, while decreases in the observed AUPP value from 340 mm (13.37 in.) to 213 mm (8.38 in.) correspond to an increase in the allowable number of ESALs from 3.7 million to 17.3 million at the research site. These results indicate that, when geogrid reinforcement is compatible with the given aggregate base material and proper construction practices are followed, statistically significant and practically important increases in pavement design life can be achieved.

The objectives of this research were 1) to examine correlations between compositional and structural properties of emulsion-treated base (ETB) layers, determine which of these factors exhibit the greatest spatial variability, and determine if significant differences exist between different test sections on a given project and 2) to investigate temporal trends in the structural properties of base materials treated with asphalt emulsion and to assess the rate at which ETB design properties are achieved. The research conducted in this study focused on testing of the ETB layer constructed on 7800 South (SR-48) in West Jordan, Utah. The research conducted in this study involved field and laboratory evaluations of spatial and temporal variability in properties of ETB. Regarding spatial results, the average modulus values of the ETB layer were unusually low for a typical stabilized base material and were in general even lower than the subgrade modulus values at this test site. All three sections had high moisture contents after compaction, with the moisture content of the ETB layer exceeding the specified optimum moisture content at many locations even before the emulsion was injected. One of the three test sections had higher percentages of reclaimed asphalt pavement and emulsion than the other two. The ETB compressive strength was very low throughout the entire year of testing, clearly demonstrating the consequences of inadequate emulsion curing associated with this project. The statistical analyses showed that higher pre-treatment moisture contents and higher amounts of binder added were associated with lower stiffness and strength, while higher wet densities were associated with higher stiffness and strength. The analyses also showed substantial variation in most response variables but comparatively low variation in predictor variables. Only four structural properties were significantly different between sections. Temporal testing was performed to monitor the properties of the ETB layer and to compare the ETB section to an adjacent untreated base course (UTBC) section. The ETB moisture content did not change significantly during the 1-year monitoring period, showing that drying of the ETB layer did not occur following placement of the hot mix asphalt surface. Furthermore, the analyses provided no evidence that the ETB layer experienced any sustained increase in strength as a result of emulsion curing; instead, the ETB modulus was shown to be greatly dependent on season, with higher ETB moisture contents and temperatures corresponding to lower ETB modulus values. Even during the winter when the ETB stiffness reached its peak, the modulus was still below the target value specified for this project. The statistical analyses indicated that the modulus values of the ETB and UTBC layers were not statistically different.

The objective of this research was to apply a previously recommended laboratory testing protocol to specific aggregate base materials that are also the subject of ongoing full-scale field testing. The scope of this research involved three aggregate base materials selected from three sites where full-scale field testing programs have been established. The first and second field sites included five different geogrid types, categorized as either biaxial or triaxial, in a singlelayer configuration, while the third site included only the triaxial geogrid type in either a singleor double-layer configuration. Geogrid-stabilized and unstabilized control specimens were evaluated using the American Association of State Highway and Transportation Officials T 307 quick shear testing protocol. Measurements of load and axial displacement were recorded and used to develop a stress-strain plot for each specimen tested. The peak axial stress, the modulus to the peak axial stress, the modulus of the elastic portion of the curve, and the modulus at 2 percent strain were then calculated. Statistical analyses were performed to investigate differences between geogridstabilized specimens and unstabilized control specimens and to investigate differences between individual geogrid products or geogrid configurations. Depending on the method of data analysis, the quick shear test results indicate that geogrid stabilization, with the effect of geogrid stabilization averaged across all of the geogrid products evaluated in this study, may or may not improve the structural quality of the aggregate base materials evaluated in this study. The results also indicate that, regardless of the method of analysis, one geogrid product or configuration may be more effective than another at improving the structural quality of a given aggregate base material as measured using the quick shear test. All results from this research are limited in their application to the aggregate base material types, geogrid products, and geogrid configurations associated with this study. Additional research is needed to compare the results of the laboratory quick shear testing obtained for this study with the structural capacity of the geogrid-stabilized and unstabilized control sections that have been constructed at corresponding full-scale field testing sites. Specifically, further research is needed to determine which method of laboratory data analysis yields the best comparisons with field test results. Finally, correlations between the results of quick shear testing and resilient modulus need to be investigated in order to incorporate the findings of the quick shear test on geogrid-stabilized base materials into mechanistic-empirical pavement design.

Loaded wheel tester (LWT) is a common testing equipment usually used to test the permanent deformation and moisture susceptibility of asphalt mixtures by applying moving wheel loads on asphalt mixture specimens. It has been widely used in the United States since 1980s and practically each Department of Transportation or highway agency owns one or more LWT(s). Compared to other testing methods for pavement materials, LWT features movable wheel loads that allow more realistic situations existing on the actual pavement to be simulated in the laboratory. Due to its potential of creating a condition of repetitive loading, the concept of using LWT for characterizing the properties of pavement materials were promoted through four innovative or modified tests in this study. (1) The first test focuses on evaluating the effect of geogrids in reinforcing pavement base courses. In this test, a base course specimen compacted in a testing box with or without geogrids reinforced was tested under cyclic loading provided by LWT. The results showed that LWT test was able to characterize the improvement of the pavement base courses with geogrids reinforcement. In addition, the results from this study were repeatable and generally in agreement with the results from another independent study conducted by the University of Kansas with similar testing method and base materials. (2) A simple and efficient abrasion test was developed for characterizing the abrasion resistance of pervious concrete utilizing LWT. According to the abrading mechanisms for pervious concrete, some modifications were made to the loading system of LWT to achieve better simulations of the spalling/raveling actions on pervious concrete pavements. By comparing the results from LWT abrasion tests to Cantabro abrasion tests, LWT abrasion test was proved effective to differentiate the abrasion resistances for various pervious concretes. (3) Two innovative LWT tests were developed for characterizing the viscoelastic and fatigue properties of asphalt mixtures in this study. In the test, asphalt beam specimens are subjected to the cyclic loads supplied by the moving wheels of LWT, and the tensile deformations of the beam specimens are measured by the LVDTs mounted on the bottom. According to the stress and strain, the parameters associated to the viscoelastic and fatigue properties of the asphalt mixture can be obtained through theoretical analyses. In order to validate the concepts associated with the above mentioned tests, corresponding conventional tests have also been conducted to the same materials in the study. According to the comparisons between the conventional and the LWT tests, the LWT tests proposed in this study provided satisfactory repeatability and efficiency.

The damaging impact of continuous utility cuts on flexible pavement performance has been shown to be a major problem for urban roads and pavement mangers due to high reconstruction and maintenance costs. Horizontal Directional Drilling (HDD) is a trenchless construction method that does not require continuous trenching. HDD pipe installation techniques can reduce reinstatement costs, shorten construction periods, and lower social costs due to reduced user traffic delays. In this thesis, a detailed field study and numerical investigations was completed to quantify pavement deterioration and polyethylene (PE) pipe performance when pipelines are installed under flexible pavements using both traditional open-cut and HDD construction methods. Two 200mm SDR-17 DIPS HDPE pipes were installed 1.5m below a flexible pavement using open-cut and HDD construction technique. A state-of-the-art instrumentation and data acquisition systems were developed to measure HDD drill rig, PE pipes and pavement responses during pipe installations and for a period of about three years afterwards. Field data from (GPR) surveys, falling weight deflectometer (FWD) tests, surface distress surveys, and ground surface elevation survey were used to evaluate pavement deterioration due to the pipeline installations. The mechanisms of ground deformations during HDD and open-cut pipe installation were numerically investigated with FLAC3D, a commercial finite difference program. A hybrid constitutive model consisting of the traditional Duncan-Chang hyperbolic model and Mohr-Coulomb perfectly plastic model was developed and implemented in FLAC3D to simulate the non-linear stress-strain and stress dependent behavior of granular materials. Field test results show that the HDD installed pipe have significantly lower construction induced strains and ring deflections when compared to the open cut-and-cover installation and the mechanism of pipe deformation differs for the two construction techniques. The two pipes performed satisfactory over the long-term monitoring period as deflections and strain levels were below acceptable limits and there was no apparent deterioration of the pipe. Pipe deflections resulting from environmental effects (freeze and thaw) were found to be more significant than those due to material creep. Furthermore, the modified Iowa’s and Plastic Pipe Institute’s (PPI) ring deflection equations were found to over estimate pipe deflection for the open-cut and HDD installed pipes by about 114 and 50 percent, respectively. Results from field tests found that the HDD installation did not results in any observable change in the condition of the pavement structure performance, while the structure and integrity of pavement section in the vicinity of the open-cut was adversely impacted by utility cut excavation. It was determined numerically that when an unsupported excavation is created within a typical flexible pavement structure, distress zones that extend laterally from the face of the excavation to a distance of approximately 80% of the depth of excavation is developed. The results of the analyses suggests that better restoration techniques are required to eliminate the adverse effect caused by the stress relief within the pavement structure during a utility cut. Furthermore, the area of potential pavement deterioration should be extended beyond the edge of the utility cut to encompass the ‘distress zones’ when determining fees to cover pavement damage and restoration costs. Results obtained from numerical simulations advanced the understanding of the mechanism, magnitude, and extent of deformation within the pavement structure during HDD pipe installation in frictional and cohesive subgrade soils. Relationship between HDD annular bore pressures and displacements have been incorporated into design Charts and Tables for use in estimating maximum allowable bore pressures for HDD installation beneath flexible pavements. Critical bore pressures that would limit ground deformations and prevent excessive pavement deformations are presented. Critical bore pressures were compared to estimated allowable bore pressures obtained from the widely used Delft Geotechnics equation. The Delft Geotechnics equation was found to over estimate allowable bore pressure for HDD installation beneath flexible pavement. HDD pipeline installations under flexible pavement were found to have significantly lower restoration costs, social costs and maintenance cost than open-cut pipeline installations.

The damaging impact of continuous utility cuts on flexible pavement performance has been shown to be a major problem for urban roads and pavement mangers due to high reconstruction and maintenance costs. Horizontal Directional Drilling (HDD) is a trenchless construction method that does not require continuous trenching. HDD pipe installation techniques can reduce reinstatement costs, shorten construction periods, and lower social costs due to reduced user traffic delays. In this thesis, a detailed field study and numerical investigations was completed to quantify pavement deterioration and polyethylene (PE) pipe performance when pipelines are installed under flexible pavements using both traditional open-cut and HDD construction methods. Two 200mm SDR-17 DIPS HDPE pipes were installed 1.5m below a flexible pavement using open-cut and HDD construction technique. A state-of-the-art instrumentation and data acquisition systems were developed to measure HDD drill rig, PE pipes and pavement responses during pipe installations and for a period of about three years afterwards. Field data from (GPR) surveys, falling weight deflectometer (FWD) tests, surface distress surveys, and ground surface elevation survey were used to evaluate pavement deterioration due to the pipeline installations. The mechanisms of ground deformations during HDD and open-cut pipe installation were numerically investigated with FLAC3D, a commercial finite difference program. A hybrid constitutive model consisting of the traditional Duncan-Chang hyperbolic model and Mohr-Coulomb perfectly plastic model was developed and implemented in FLAC3D to simulate the non-linear stress-strain and stress dependent behavior of granular materials. Field test results show that the HDD installed pipe have significantly lower construction induced strains and ring deflections when compared to the open cut-and-cover installation and the mechanism of pipe deformation differs for the two construction techniques. The two pipes performed satisfactory over the long-term monitoring period as deflections and strain levels were below acceptable limits and there was no apparent deterioration of the pipe. Pipe deflections resulting from environmental effects (freeze and thaw) were found to be more significant than those due to material creep. Furthermore, the modified Iowa’s and Plastic Pipe Institute’s (PPI) ring deflection equations were found to over estimate pipe deflection for the open-cut and HDD installed pipes by about 114 and 50 percent, respectively. Results from field tests found that the HDD installation did not results in any observable change in the condition of the pavement structure performance, while the structure and integrity of pavement section in the vicinity of the open-cut was adversely impacted by utility cut excavation. It was determined numerically that when an unsupported excavation is created within a typical flexible pavement structure, distress zones that extend laterally from the face of the excavation to a distance of approximately 80% of the depth of excavation is developed. The results of the analyses suggests that better restoration techniques are required to eliminate the adverse effect caused by the stress relief within the pavement structure during a utility cut. Furthermore, the area of potential pavement deterioration should be extended beyond the edge of the utility cut to encompass the ‘distress zones’ when determining fees to cover pavement damage and restoration costs. Results obtained from numerical simulations advanced the understanding of the mechanism, magnitude, and extent of deformation within the pavement structure during HDD pipe installation in frictional and cohesive subgrade soils. Relationship between HDD annular bore pressures and displacements have been incorporated into design Charts and Tables for use in estimating maximum allowable bore pressures for HDD installation beneath flexible pavements. Critical bore pressures that would limit ground deformations and prevent excessive pavement deformations are presented. Critical bore pressures were compared to estimated allowable bore pressures obtained from the widely used Delft Geotechnics equation. The Delft Geotechnics equation was found to over estimate allowable bore pressure for HDD installation beneath flexible pavement. HDD pipeline installations under flexible pavement were found to have significantly lower restoration costs, social costs and maintenance cost than open-cut pipeline installations.

Fatigue cracking is one primary distress in asphalt pavements, dominant especially in later years of service. Prediction of mixture fatigue resistance is critical for various applications, e.g., pavement design and preventative maintenance. The goal of this work was to develop a tool for prediction of binder aging level and mixture fatigue life in pavement from unaged binder/mixture properties. To fulfill this goal, binder oxidation during the early fast-rate period must be understood. In addition, a better hourly air temperature model is required to provide accurate input for the pavement temperature prediction model. Furthermore, a user-friendly software needs to be developed to incorporate these findings. Experiments were conducted to study the carbonyl group formation in one unmodified binder (SEM 64-22) and one polymer-modified binder (SEM 70-22), aged at five elevated temperatures. Data of SEM 64-22, especially at low temperatures, showed support for a parallel-reaction model, one first order reaction and one zero order reaction. The model did not fit data of SEM 70-22. The polymer modification of SEM 70-22 might be responsible for this discrepancy. Nonetheless, more data are required to draw a conclusion. Binder oxidation rate is highly temperature dependent. Hourly air temperature data are required as input for the pavement temperature prediction model. Herein a new pattern-based air temperature model was developed to estimate hourly data from daily data. The pattern is obtained from time series analysis of measured data. The new model yields consistently better results than the conventional sinusoidal model. The pavement aging and fatigue analysis (PAFA) software developed herein synthesizes new findings from this work and constant-rate binder oxidation and hardening kinetics and calibrated mechanistic approach with surface energy (CMSE) fatigue analysis algorithm from literature. Input data include reaction kinetics parameters, mixture test results, and pavement temperature. Carbonyl area growth, dynamic shear rheometer (DSR) function hardening, and mixture fatigue life decline are predicted as function of time. Results are plotted and saved in spreadsheets.

Within the transit industry it is well known that transit buses have the potential to operate at weights that exceed vehicle weight limits. However, few attempts have been made to date to determine how often this occurs and to what degree. This research characterizes the current transit industry with respect to the regulatory environment, factors that have affected the weight of modern day transit buses, and methods for accommodating transit buses in pavement design. This research then develops and applies a methodology for calculating the in-service weights of standard 40-ft. transit buses using a combination of passenger characteristic data, transit bus curb weight data, and transit ridership data. The findings of this research suggest that the transit bus industry is in a state of competing interests. Weight estimates developed in this research identify that current transit bus models are unable to comply with vehicle weight limits in most jurisdictions even with no passengers on board. Further, these estimates indicate that transit buses have a significant impact on pavements – comparable to those of fully-loaded, five-axle semi-trucks on a per vehicle basis. To date this issue has been addressed in the Canadian Prairie Region by indefinitely granting transit buses overweight permits. However, based on the current state of the transit industry there is little incentive for transit agencies to operate lightweight transit buses and little incentive for transit bus manufacturers to produce lightweight transit buses in order to address pavement and regulatory concerns. Consequently, transit bus axle weight issues in the Canadian Prairie Region are expected to continue in the foreseeable future. / October 2015

Asphalt mixtures are complex composites that comprise aggregate, asphalt binder, and air. Several research studies have shown that the mechanical behavior of the asphalt mixture is strongly influenced by the matrix, i.e. the asphalt binder. Therefore, accurate constitutive models for the asphalt binders are critical to ensure accurate performance predictions at a material and structural level. However, researchers who use computational methods to model the micromechanics of asphalt mixtures typically assume that (i) asphalt binders behave linearly in shear, and (ii) either bulk modulus or Poisson’s ratio of asphalt binders is not time dependent. This research develops an approach to measure and model the shear and bulk behavior of asphalt binders at intermediate temperatures. First, this research presents the findings from a systematic investigation into the nature of the linear and nonlinear response of asphalt binders subjected to shear using a Dynamic Shear Rheometer (DSR). The DSR test results showed that under certain conditions a compressive normal force was generated in an axially constrained specimen subjected to cyclic torque histories. This normal force could not be solely attributed to the Poynting effect and was also related to the tendency of the asphalt binder to dilate when subjected to shear loads. The generated normal force changed the state of stress and interacted with the shear behavior of asphalt binder. This effect was considered to be an “interaction nonlinearity” or “three dimensional effect”. A constitutive model was identified to accommodate this effect. The model was successfully validated for several different loading histories. Finally, this study investigated the time-dependence of the bulk modulus of asphalt binders. To this end, poker-chip geometries with high aspect ratios were used. The boundary value problem for the poker-chip geometry under step displacement loading was solved to determine the bulk modulus and Poisson’s ratio of asphalt binders as a function of time. The findings from this research not only improve the understanding of asphaltic materials behavior, but also provide tools required to accurately predict pavement performance. / text