Stacy G. Williams

EXAMINATION OF OPERATOR VARIABILITY FOR SELECTED METHODS FOR MEASURING BULK SPECIFIC GRAVITY OF HOT-MIX ASPHALT CONCRETE

The ability of different operators to obtain similar results when performing laboratory tests on the same material is vital for producing accurate testing results. By conducting trials in triplicate for each of three different testing methods, a measurement of the bulk specific gravity (Gmb) of compacted hot-mix asphalt concrete (HMAC) cores was obtained. An analysis of the variability between operators was investigated using a total of almost 1,300 test results, using HMAC sampled from six projects in Arkansas. Three methods were used to determine the bulk specific gravity of compacted HMAC samples, including saturated surface dry (SSD) (as per AASHTO T166), height and diameter (as per AASHTO T269), and vacuum sealing (using the Corelok vacuum sealing device). In almost all cases, Gmb values determined using the height and diameter method were statistically different from those determined using the SSD and Corelok methods; further, statistical differences were noted in paired analyses between the SSD and Corelok methods. The Corelok method exhibited a lower degree of variability than the other two methods used, based on the standard deviation of test results obtained by different operators. In direct comparison with the SSD method, the Corelok exhibited a lower variability (standard deviation) in 81 percent of the cases. Overall, the Corelok method appears to offer a viable alternative for determining the bulk specific gravity of compacted HMAC. However, agencies seeking to use the Corelok must consider the effect of an apparent shift in Gmb values obtained on resulting HMAC volumetric and compaction properties.

Methods for Evaluating Longitudinal Joint Quality in Asphalt Pavements

Longitudinal joint quality is essential to the successful performance of hot-mix asphalt (HMA) pavements. Longitudinal joints have received a considerable amount of attention recently, because many state agencies are moving toward implementation of longitudinal joint specifications. Most measures of joint quality are based on density determinations. However, distress at the joint is caused by the ability of air and water to enter the pavement structure, an action that is directly related to the permeability of the joint. In other words, density alone may not be sufficient to describe the quality of a longitudinal joint. The objective of this study was to identify the most appropriate test method or methods for describing HMA longitudinal joint quality. Three projects were selected for the study, and four testing stations were identified for each project. At each station, a number of test methods were performed at the longitudinal joint and to either side of the joint to assess the ability of each method to properly discriminate between levels of joint quality. Multiple measures of density, permeability, and gradation were obtained at each station. Overall, the methods providing the most accurate predictions of joint quality and the greatest level of discrimination were the vacuum sealing and saturated surface-dry (SSD) methods for determining the bulk specific gravity of field cores. Although measures of permeability and infiltration showed promise, it was recommended that the vacuum sealing and SSD methods for the determination of field core density be considered in the evaluation of longitudinal joint quality.

Precision of AASHTO TP62-07 for Use in Mechanistic—Empirical Pavement Design Guide for Flexible Pavements

In the Mechanistic–Empirical Pavement Design Guide (MEPDG), one of the most important material inputs for designing flexible pavements is the dynamic modulus (E*). Currently, AASHTO has provided a provisional test specification for measuring the E* of hot-mix asphalt mixtures, AASHTO TP62-07 [Standard Method of Test for Determining Dynamic Modulus of Hot-Mix Asphalt (HMA)]. Under AASHTO TP62-07, recommendations are provided for sample preparation–compaction, measurement system, testing conditions–procedures, and calculations. Unfortunately, information regarding the general precision of the test method does not exist. With many different test machines capable of conducting AASHTO TP62-07 and the importance of E* in flexible pavement design, an understanding of the general precision of the test method and its influence on the MEPDG is needed. This paper describes the results of a round-robin study conducted to address the need for a precision-type statement for AASHTO TP62-07. Seven laboratories, all having equipment compliant with the specifications in AASHTO TP62-07, were asked to laboratory age, compact, prepare, and test six HMA test specimens according to the provisions in AASHTO TP62-07. A statistical analysis was conducted on the test data in a precision statement environment. On the basis of the test results, an initial precision statement for AASHTO TP62-07 was generated and presented. The test results for the seven different laboratories were also used as inputs in the MEPDG to evaluate the relative differences in predicted pavement distresses. Results from the precision evaluation were used to recommend modifications to AASHTO TP62-07 to promote greater precision in different testing laboratories and the results were reevaluated in the MEPDG.

ESTABLISHING VARIABILITY FOR HOT-MIX ASPHALT CONSTRUCTION IN ARKANSAS

The Arkansas quality control/quality assurance (QC/QA) specification for hot-mix asphalt (HMA) construction was initially developed from data and experience gained with the Marshall method of mix design. The implementation of Superpave®, coupled with the initiation of the largest asphalt construction program in the state’s history, resulted in questions relating to the suitability of the QC/QA specification. Research was undertaken to revise the existing program or create a new specification. The first step in the effort was to establish typical HMA construction variability in terms of the pay factors used in Arkansas: air voids, voids in the mineral aggregate, binder content, and field density. Six projects were randomly sampled and tested by three operators: the contractor, the agency, and the research team. Six specimens were prepared by each operator at each of four testing times within each project. The results of the testing showed that the six projects bracketed a wide variation in mixture quality as defined by the HMA properties. The testing data were normally distributed, validating a typical assumption regarding construction. Analyses showed that the testing variability among the three operators could be absorbed into the overall variability of test properties instead of being considered a distinct factor. Three levels of HMA quality were identified by the mean and standard deviation of the test data. These quality levels are expected to form the basis for any revision to the existing QC/QA system or for the development of a new QC/QA system for Arkansas HMA construction.

Comparison of the Superpave Gyratory and Proctor Compaction Methods for the Design of Roller-Compacted Concrete Pavements

Roller-compacted concrete (RCC) is gaining popularity as a paving alternative for various roadway applications. RCC is a zero-slump concrete mixture that does not require forms, reinforcing, jointing, or finishing. This type of material combines the advantages of conventional concrete pavements with the ease of construction of an asphalt pavement and can be opened to traffic more quickly than conventional concrete pavements. The Superpave® gyratory compactor was investigated as an alternative to Proctor compaction (ASTM D1557) for the design of RCC paving mixtures. Seventeen mix designs that encompassed a range of aggregate sources, gradation types, and natural sand contents were generated. In general, the gyratory method did not produce designs equivalent to the Proctor method, in that the gyratory method resulted in higher densities and lower optimum moisture contents. Gyratory compaction tended to generate a maximum, or terminal, achievable moisture content, regardless of the amount of moisture mixed in the specimen. The corresponding densities continued to increase rather than display the parabolic relationship typical of Proctor compaction. The highest densities by both compaction methods tended to be achieved for blend gradations that most closely followed the maximum density line and included the greatest percentages of natural sand.

Bulk Specific Gravity Measurements of 25.0-mm and 37.5-mm Coarse-Graded Superpave Mixes

The measurement of specific gravity for hot-mix asphalt (HMA) is critical to almost every activity pertaining to asphalt mixtures. During design and construction, the bulk specific gravity (Gmb) of HMA mixes is used to calculate many properties that indicate pavement quality. Thus, it is important to measure this property accurately and precisely. This study evaluated four methods for measuring Gmb, including the traditional saturated surface dry (SSD) method, the CoreLok method, the height-diameter method, and the CoreReader method. Four aggregate sources were used to create mix designs for 25.0-mm and 37.5-mm nominal maximum aggregate sizes (NMASs). In addition, the level of compactive effort was varied to assess the comparisons for a range of air voids. Results indicate that the traditional method exhibited the lowest levels of variability. Strong correlations were developed to relate the test methods on the basis of the large NMASs. The strongest and most practical relationship was developed between the SSD and CoreLok methods. These models can be used to assess the effects of incorporating new test methods for both design and construction procedures. They can also be used to normalize data from different test methods, should alternative methods be desired. Although some recommend the use of an alternative test method (such as CoreLok), the studies on which these recommendations are based did not include large NMAS mixes (25.0 mm and 37.5 mm). Although alternative methods do possess significant advantages, the results of this study do not support the elimination of the traditional Gmb test method.

Sample Size Requirements for Field Permeability Measurements of Hot-Mix Asphalt Pavements

The permeability of coarse-graded asphalt mixtures has been of great concern in recent years. Asphalt mixes that are permeable are susceptible to a number of distresses, such as moisture damage, raveling, cracking, and binder oxidation. In this study, seven sites were mapped according to the vacuum permeameter method to determine the location and distribution of permeable voids. The variability of the sites was evaluated to determine an appropriate sample size for permeability testing. Relatively impermeable pavements were consistent; pavements with moderate or high permeability were more variable. Relationships between sample size, reliability, and testing discrimination are presented. Based on the range of standard deviations measured in this project, a minimum sample size of 10 is recommended for pavements with marginal densities or variable consistency. On the basis of the results of this study, permeability test results are highly dependent on the placement of the permeameter during testing, the effective testing area of the permeameter, and the variability of the pavement. Field permeability testing may have value as a forensic tool, but the large sample size required is not conducive to standard quality control procedures.

Density, Permeability, Infiltration, and Absorption Used to Assess Quality of Hot-Mix Asphalt Longitudinal Joints

Longitudinal joint quality is critical to the successful performance of hot-mix asphalt pavements. The implementation of longitudinal joint quality specifications is gaining momentum in many states, with density being the most common quality parameter. However, distress at the joint is caused by the ability of air and water to enter the pavement structure, an ability that may be more compatible with a measure of permeability. The objective of this study was to provide guidance for the implementation of joint quality specifications by compiling information that more completely describes the relationships of density, absorption, permeability, and infiltration. On two resurfacing projects, eight joint construction techniques were used, and the ability of each to improve joint quality was assessed. In addition, joint quality was evaluated with regard to proximity to the joint. In general, the joint heater, joint stabilizer, and notched wedge construction methods were the best performers in regard to density and permeability. Traditional rolling methods exhibited lesser quality, indicating that additional measures may be necessary to generate acceptable levels of joint quality. For the most appropriate measure of quality, permeability and infiltration adequately discriminated between varying levels of joint quality, as did nuclear density. Reasonable relationships were developed between density, permeability, infiltration, and absorption. Thus, density measurements can be used for assessing joint quality and simultaneously protecting against excessive levels of permeability.

Critical Factors Affecting Field Determination of Hot-Mix Asphalt Density Using Nonnuclear Devices

Nonnuclear methods for the measurement of hot-mix asphalt (HMA) density offer the ability to take numerous density readings in a short time period, without the need for intensive licensing, training, and maintenance efforts common to nuclear gauges. The Pavement Quality Indicator (PQI) and the PaveTracker use electrical impedance to estimate density. Early models of these gauges were deemed inadequate for quality control and quality assurance testing, but improvements have been made to each. In this project, a ruggedness study was performed to evaluate the effects of gauge model, temperature, moisture, use of sand to fill surface voids, gauge orientation, number of replicate measurements, and gauge placement on two nonnuclear gauges. Significant factors arising from the ruggedness study were further investigated. Two 12.5-mm mixes and one 37.5-mm mix were evaluated, using 42 test locations. Gauge orientation was significant, suggesting that the test method protocol should specify gauge orientation with respect to the direction of paving. Four replicate measurements were deemed acceptable for producing a single reported result, and all sand and debris should be thoroughly brushed from the surface before testing. Moisture significantly affected the measured densities in some cases, most notably for the 37.5-mm test locations and for the PaveTracker device.

EXAMINATION OF GAMMA-RAY METHOD FOR MEASURING BULK SPECIFIC GRAVITY OF HOT-MIX ASPHALT CONCRETE

Much attention has been given to the determination of bulk specific gravity (Gmb) of compacted hot-mix asphalt (HMA) specimens. For relatively open-graded HMA mixtures, such as those produced by Superpave® and stone matrix asphalt design methods, perceived potential shortcomings in traditional test methods may be pronounced. Recently, a new device was introduced for measuring Gmb that features the use of gamma-ray technology. To evaluate the efficacy of the device, 221 compacted HMA specimens were tested with three test methods: saturated surface dry (SSD) (AASHTO T-166), vacuum-sealing (Corelok), and gamma-ray technology (CoreReader). Each specimen was tested three times in each of the three methods. The specimens represented field-sampled and laboratory-prepared mixtures, 12.5-mm and 25-mm nominal maximum aggregate sizes, and a variety of aggregates typically used in Arkansas. Test results indicated that statistically significant differences exist between mean Gmb-values generated by each of the three testing methods. In general, the vacuum-sealing method yielded the lowest Gmb-values, followed (in order of low to high) by the gamma-ray and SSD methods. Significant differences in the variability of test results were not observed between test methods. In ease of testing and testing variability, the CoreReader gamma-ray device does not appear to offer significant advantages for determining the bulk specific gravity of compacted HMA cores. Further, departments of transportation wishing to replace or allow alternatives to traditional SSD-based tests should consider the impact of test results on existing volumetric property specifications.