The responses of typical flexible and rigid pavements to the dynamic loads imposed by the rolling dynamic deflectometer (RDD) were investigated. The RDD is a recently developed nondestructive testing vehicle that applies a steady-state harmonic force while continuously moving along the pavement. To find the pavement response to moving harmonic loads, a numerical method using the elastodynamics solution for stationary impulsive loads applied to layered media and the superposition of the effect of sequential loading impulses at different stations (spaced according to the advance velocity of the load) was implemented. The effects of RDD velocity, loading frequency, depth to bedrock, and layer stiffness on the deflection basins were investigated in the longitudinal and transverse directions. The results show that for the vehicle velocities and loading frequencies that will be used in conjunction with the RDD, dynamic effects caused by the motion of the vehicle are negligible, and the deflection basins at any instant of time essentially are the same as those that would be obtained with a stationary static or harmonic load, with the advantage to RDD testing being that the loading magnitude and loading frequency can be changed easily and that profiling is continuous along the pavement.

The classical Westergaard idealization of the rigid pavement system assumes that the slab-on-grade can be modeled as a thin elastic plate on a bed of springs. Implicit in this assumption is the Kirchhoff (or thin) plate theory, which ignores transverse shear stresses. This same assumption has been carried over into the development of a number of two-dimensional finite-element models developed for rigid pavements. In this paper, modern isoparametric two-dimensional and three-dimensional finite-element formulations are employed to show that the Kirchhoff assumptions are not strictly valid for the edge loading case in rigid pavements. For slabs where the span-to-depth ratio is less than 100, classical Kirchhoff assumptions, adopted by Westergaard, lead to errors in predicting edge stresses. The maximum edge stress is, in fact, less than that predicted by Westergaard. Furthermore, the maximum edge stress does not occur at the edge of the slab, but at some finite distance from the edge. For most practical rigid pavement systems, the maximum edge stress will occur at a finite distance away from the edge of the slab and will be approximately 10% less than that predicted by the Westergaard theory.

A novel application of the principles of fracture mechanics combined with multilayered elastic theory is used to obtain the design stresses midway between the transverse joints at the bottom of the longitudinal edge of rigid pavements and for corner loading of slabs without dowels. The thermal curl stresses and deflection for slabs that are curled up, with and without vehicular loading, are also determined. The basic idea is to treat a joint with no load transfer, as a crack in a continuously supported plate on an elastic foundation. The deflection and stresses caused by the joint are determined by fracture mechanics, and those for the continuous pavement are determined by the CHEVRON multilayered elastic computer program. The total deflection and stresses are then obtained by superposition. The deflection and stresses so determined are compared with the Westergaard theoretical solutions, the results of finite-element programs, and experimental data from the American Association of State Highway Officials Road Test, showing good agreement.

This paper presents a procedure for characterizing granular base and subbase materials to provide the properties needed by the mechanistic methods for pavement design. It includes repetitive testing under different confining pressures and axial stresses. It is different from the resilient modulus test. Each specimen is tested at a different confining pressure for 10,000–100,000 load repetitions. Both the resilient and the permanent responses are collected. The sample compaction, the setup for measuring vertical and horizontal deformations, and the repetitive testing are described. Test results at different confining pressures and axial stresses show that (1) the material exhibits hysteretic behavior, (2) the material response is nonlinear [after a large number of load repetitions (of the order of 10,000) the material may be considered as elastic], (3) the material dilates in the resilient mode; (4) the resilient behavior can be modeled reasonably using the extended bulk modulus equation or the hypoelastic model, and (5) the variability in the material response indicates that the specimen size is, at best, marginal for the aggregate gradation used in the study.

The fatigue life of pavements modeled as Winkler beams, with irregular profile, and traversed by vehicles modeled as nonlinear dynamic systems, is investigated in this article. Three computational methods are used to integrate the road equilibrium equation: the time domain superposition (Duhamel's integral), the frequency superposition based on a reference frame moving with the vehicle, and a very simple pseudodynamic scheme based on the static influence function in conjunction with the fluctuating loads. It will be shown that the frequency superposition method is significantly more efficient than time superposition, and that the simplified pseudodynamic method could be used as a reasonable first approximation. A vehicle model is used in conjunction with a profile model in order to obtain a sample force record with which to load the pavement. The response of the road is analyzed, and its fatigue life estimated, for different values of the parameters that describe the pavements. It will be shown that pavement thickness and modulus have a tremendous influence on fatigue life.

A great deal of research has been carried out over the past few decades in seeking explanatory variables to understand various human judgmental indices of road quality. The essential question is, What is a rater sensing in a dynamical vibration environment while in a vehicle moving on a road? And what does a rating given by a rater represent? This is an interesting operational problem dealing with a complex system, a problem embracing road profile, vehicle dynamic response, and human kinesthetic reaction to a vibration environment. It is suggested that the jerk experienced by raters is one of the most important variables in explaining human ratings of roads. A comprehensive investigation is carried out to show that the jerk plays a vital role in explaining the serviceability of roads. Comments are given on other road indices. Dependence of human rating on the speed of moving vehicles is also addressed.

Two existing models of automobile crashes are extended and compared to improve the evaluation of the performance of new, vehicle-based technologies for crash avoidance. The crashes model is based on knowledge of the frequencies and severities of common crash scenarios. The fault tree model is based on knowledge of the frequencies of and relationships among accident causation events. New technology countermeasures, such as drowsy-driver warning and center-mounted brake lights, can be studied individually and in combination using either model. The effectiveness of the countermeasure technologies is measured as the reduction in the frequency and cumulative harm of crashes. For some technologies, it is more straightforward to estimate the effect of technology to mitigate a well-defined accident scenario, as in the crashes model. For other prospective technologies, it also is possible to measure the influence of technology on a well-defined initiating event, as in the fault tree model.

Given the increasing use of double left-turn lanes and the lack of methods to design and analyze them, there exists an urgent need to develop such methods. This paper describes work completed at North Carolina A&T State University that expands the scope of previous work done in the area of double left-turn lanes and improves the scientific basis of that work. Double left-turn data from 30 intersections in the city of Greensboro, North Carolina, were analyzed with respect to relationships among performance and geometric characteristics. Results indicate that the overall f_LT of 0.92, recommended by the Highway Capacity Manual, appears reasonable for the inside lane. Values of 0.86–0.88 are more appropriate for outside lane adjustments. The variables causing the most significant changes in performance of double left-turn lanes were traffic related, and not related to geometry. The presence or absence of through traffic and the absence of opposing left-turn flow were key causes of performance changes.

The selection of an appropriate alignment for a proposed highway is determined largely by relating topographic, urban, and environmental features to geometric design controls. Typically, aerial photographs and topographic, geologic, and soil maps are reviewed. In this paper, a geographic information system (GIS) platform that incorporates the main coverages needed for evaluating route alignments is described. Using the GIS and a geographically referenced database, a decision-aid tool for multicriteria evaluation of route alignments is developed. Possible alignments are evaluated based on community disruption and environmental, geotechnical, and geometric design criteria. The developed decision-aid tool integrates slope stability and roadway design packages and specifically written codes with GIS packages ARC/INFO and ArcView, the latter acting as the system engine and interface. A case study is presented that applies the developed platform to the testing of potential alignments for a proposed 12 km highway to the south of the city of Beirut, Lebanon. Results of the case study demonstrated the advantages of the decision-aid tool and highlighted its potential in providing a quick, multicriteria screening evaluation of possible route alignments.

Simulation of urban traffic noise along with a mathematical model of stop-and-go traffic noise for the road network in the city center is presented in this paper. The road network of the central part of Bangkok surrounded by the Cordon line of Rajchadapisek Ring Road was used as the study area. The analysis data consisted of traffic characteristics and its noise levels, which were collected from the study area. The geometrical dimensions of road section at each data collection location also were measured. Characteristics of traffic noise from different types of vehicles were analyzed, including tuk-tuks (motorized tricycle taxis) and motorcycles, which are considered as popular vehicle types in Bangkok. Characteristics of traffic noise levels and data on other traffic noise parameters were used to analyze and subsequently construct the stop-and-go traffic noise simulation model. Finally, the resultant stop-and-go traffic noise models, which are separated into acceleration and deceleration lane models and their statistical goodness-of-fit tests, are presented.

Preemption techniques are designed to provide preferential treatments to buses at signalized intersections. A preemption strategy, if properly designed, can provide continuous green phases to buses at successive intersections, thereby reducing delays and travel times along the bus-street. It can also potentially increase delays and travel times along the cross street. A procedure for economic evaluation of preemption projects is presented in this paper, followed by a case study demonstration. It is a variation of benefit-cost analysis that provides an estimate of the maximum investment that can be justified in a signal preemption project by way of savings in delay, fuel consumption, and emissions. The TRAF-NETSIM model was used to assess the operational consequences of signal preemption. The procedure is found viable and the case study results appear to be reasonable, and sensitive to changes in interest rates and project lives.