A common laboratory fatigue test for life estimation of pavements uses a repeated load flexural device with nonlayered beam specimens. As a result of such factors as traffic wander, and also due to different crack growth rates between such beam specimens and layered media, highway pavements have been found to sustain from 10 to 100 times the number of load applications estimated by this type of laboratory test. This study focuses on identifying the differences in crack propagation between beam specimens and layered pavements. A 2D finite-element program with specialized fracture mechanics capabilities is used in the analysis. An isotropic linear elastic material is assumed. Noncohesive cracks are introduced in the mesh as geometrical discontinuities. Cracks are propagated through the material under cyclic loading. Stress intensity factors for mixed mode are calculated for each crack length. Two different configurations are analyzed: laboratory beam specimens and layered pavements. It will be shown that predicted crack growth rates in pavements for the same maximum longitudinal strain are much lower than in beam specimens, which is consistent with laboratory and field observations. Two different crack growth rates are identified in pavements. The first is associated with tensile stresses at the bottom of the surface course. The second one is caused by tensile stresses at the top of the surface course, when the load is ahead of or behind the crack. The growth rate for surface cracks (growing downward) is much smaller than the internal crack growth rate (growing upward).

Properly specified pavement deterioration models are an important input for the efficient management of pavements, the allocation of cost responsibilities to various vehicle classes for their use of the highway system, and the design of pavement structures. However, most empirical deterioration progression models developed to date have had limited success. This paper is concerned with the development of an empirical rutting progression model using experimental data. The data used in this paper comprise an unbalanced panel data set with more than 14,000 observations taken from the AASHO Road Test. The salient features of the model specification are (1) the model eschews conventional (predefined) axle load equivalencies and structural numbers in favor of relationships determined entirely by the data itself; (2) a thawing index variable has been incorporated to capture the effects of the environmental factors in the AASHO Road Test; and (3) the model predicts incremental changes in rut depth, which is particularly advantageous in a pavement management context. The specified model is nonlinear in the variables and the parameters and is estimated using both fixed-effects and random-effects specifications to account for unobserved heterogeneity. The estimation results show that the model replicates the pavement behavior well, that the inclusion of an environmental variable is important to avoid biases in other parameters, and that the size of the unobserved heterogeneity is significant. It is also found that interactions betwen some parameters in the nonlinear specification leads to significant differences between parameter estimates among the two wheel paths rutting models.

A reliability based mix design and analysis system for thermal cracking in asphalt concrete pavements is described and demonstrated in this paper. Asphalt-aggregate mixture performance in terms of low temperature cracking and thermal fatigue is assessed independently for each type of cracking in the specified environment. Reliability factors are introduced into this assessment to account for the level of risk assumed by the engineer and the inherent variability in the estimates of mixture resistance to induced thermal stresses and the environmental demand in terms of conditions promoting either type of thermal cracking. Mixture resistance is measured in two laboratory tests, and environmental demand is estimated from predicted surface pavement temperatures. The paper concludes with an example of the analysis described, comparing unmodified and crumb-rubber modified mixture performance at different levels of reliability in a high elevation desert climate. The system presented is recommended as a mix design and analysis tool for evaluating unmodified and modified asphalt-aggregate mixture performance in harsh climates where thermal cracking is of concern.

This research develops a model of prework trip-making and home departure time choice. A nested logit model was found to be an appropriate approach. Data from the New York Metropolitan Statistical Area, collected as a part of the 1995 Nationwide Personal Transportation Survey (NPTS), were used for empirical estimation of model parameters. The model results confirmed several behavioral aspects associated with prework trip-making and home departure time choice. The paper presents a discussion on possible policy implications that can be inferred from the results. Future potential research questions and data requirements for an elaborate modeling approach are also briefly discussed.

The 1985 European Union Directive on Environmental Impact Assessment was transported into Irish law in 1988. The resulting complex environmental impact statements often contain large volumes of information on relevant impacts, yet make relatively little effort to organize this information into a decision-assisting framework that enables the relative merits of the options to be comprehensively compared. This paper describes a practical application of the ELECTRE III decision-aid system to a preliminary environmental appraisal carried out recently on the Port Access and Eastern Relief Route (PAERR), a new motorway proposed for Dublin City. The case study illustrates how the ELECTRE III framework takes complex information from the environmental appraisal and uses it to rank the various project alternatives considered. The paper describes new techniques put forward by the writers for use within the ELECTRE III model. A weighting system previously developed by the writers is used for estimating the relative importance of criteria. Also, the writers' own methods are used for estimating indifference, preference, and veto thresholds. These estimates allow the uncertainties inherent in each set of criterion valuations to be incorporated directly into the decision model. Both these techniques make the model more readily applicable to the decision problem under examination within this paper.

One of the key elements in the evolution of traffic flow theory has been the development of the car following theory. Applications of this research can be seen in traffic simulation models, such as CORSIM, that use car following models to simulate traffic flow and predict operational performance characteristics on arterial roadways. Researchers at Louisiana State University have recently developed a technique using the Global Positioning System to record and analyze car following behavior. In this study, these procedures were used to collect vehicle motion data and compare aspects of driver behavior recorded in the field to those predicted by the General Motors car following models. These comparisons were also used to develop numeric values of driver sensitivity and time lag from field data.

Traffic signal warrants documented in the Manual on Uniform Traffic Control Devices neglects the effect of daily and hourly volume variations in traffic. Furthermore, current practice cannot be used to prioritize signal installations among multiple intersections that meet the warrants. Finally, there is no guidance in the Manual on Uniform Traffic Control Devices on how the warrants can be applied to projected traffic conditions when hourly counts are not available. This paper describes a probabilistic approach to implementing traffic signal warrants. The proposed approach used Monte Carlo simulation to generate, on the basis of variable hourly traffic counts, the probability that any particular signal warrant will be met at an intersection. The simulation approach was extended to develop planning level guidelines for signal installation using predefined synthetic traffic flow profiles and the average daily traffic estimates for the intersecting streets. The proposed guidelines were compared with current planning guidelines used in Texas and California. The comparison results indicate that the Texas guidelines tend to underestimate the need for traffic signal installation, whereas the California guidelines tend to overestimate the need.

This paper reports on the findings of pedestrian flow characteristics for different types of walking facilities in Hong Kong. Surveys were conducted during peak hours at indoor and outdoor walkways, signalized crosswalks with and without midblock, signalized and nonsignalized crosswalks leading to light-rail transit stations, and five types of walking facilities in railway stations. The collected data were used to develop speed/flow relationships and to calibrate the travel time function for each type of pedestrian facility. Pedestrian flow characteristics on various walking facilities are also compared. The collected data and the established relationships could be used as a basis for the development of pedestrian simulation models for Hong Kong and/or other Asian cities with similar pedestrian characteristics.

This paper shows how principal components analysis (PCA) can be used to analyze a large set of real trajectories of vehicles on a bend. PCA gives an opportunity to reduce the mathematical and numerical operations necessary to extract real values of traffic parameters (such as curvature, centrifugal acceleration, and the rate of change of this acceleration) from the set of trajectories. Also profiles of these parameters can be obtained. Applied to an exemplary set of real trajectories, PCA enabled finding that maximum curvature and rate of change of centrifugal acceleration are square functions of vehicle velocity, and maximum acceleration is its linear function.

The development of public housing has major traffic impacts on road networks, particularly in already congested cities. Hong Kong, being one of the most congested cities in the world, has a significant amount of public housing in town. Severe traffic congestion occurs in the adjacent road network if the forecast of traffic generation is not accurate. This technical note develops a multilinear regression model to forecast traffic generation for large-scale, high-density, multistory public residential housing estates in Hong Kong. Manual traffic counts were conducted at 36 housing development sites throughout Hong Kong. Regression analysis was undertaken for model evaluation. Many variables, such as the number of apartments, population, gross floor area, parking spaces, and accessibility, were included in the analysis, and the resulting model is both qualitatively and quantitatively reasonable.