The Alaska Department of Transportation and Public Facilities (ADOTPF) is presently evaluating the use of recycled rubber in hot mix pavement applications. The benefits of adding rubber to the mix include increased skid resistance under icy conditions, improved flexibility and crack resistance, elimination of a solid waste, and reduced traffic noise. The major disadvantage of these rubber‐modified mixes is their high cost in relation to conventional asphaltic concrete pavements. A comparison of the economics of the rubber‐modified system with that of the conventional pavement shows that the rubber‐modified surfacing is cost‐effective. This conclusion is based on an analysis of life‐cycle costs, but does not include potential intangible benefits of the rubbermodified system, such as increased skid resistance and noise reduction.

This study evaluates the special features of premium design guidelines—features not considered in the AASHO flexible and rigid pavement design procedures. The significance of these features—fifteen in all for four pavement types—were investigated by evaluating pavement performance and design reliability. The researchers used the VESYS III program to evaluate features of the flexible pavements and an algorithm developed in a companion paper for the other three types (composite, jointed plain concrete, and continuously reinforced concrete). This computer program, Reliability Analysis and Performance of Pavements I (RAPP‐I), employs Monte Carlo simulation techniques to treat all of the design variables probabilistically. The effectiveness of each feature is evaluated by comparing the performance or expected life and reliability of typical pavement sections with and without a premium feature. The features beneficial in virtually all the pavements include adequate drainage, a nonerodable base, or a monolithic/tied shoulder. The rational method of selecting bitumen grade appears promising in flexible and composite pavements. The joint spacing‐recommendation of premium design procedure contributes to the longevity of jointed plain concrete pavements, as does the use of dowel bars. The significance of quality control and quality assurance of concrete is clearly demonstrated in the study.

A simulation model to calculate the reliability/performance of pavements is developed. The computer program, Reliability Analysis and Performance of Pavements (RAPP‐I), employs Monte Carlo simulation techniques to solve the design equations (e.g., AASHTO, Premium) in which all of the design variables are assumed to be probabilistic and normally distributed. RAPP‐I, in conjunction with the respective design model, calculates the present serviceability index (PSI) of pavements in one‐year increments. In addition to uncertainties attributable to the design factors, errors due to idealization of the model are included in the PSI calculation. By comparing the computed PSI with the terminal PSI, assuming that both are normally distributed, standard probabilistic techniques are employed for calculating pavement reliability. An expression for calculating the “expected” life of pavements, employing the reliabilities at various ages, is proposed. The expected life is found to be a convenient measure for comparing the performance of various pavement design features.

The problem of scheduling maintenance for pavements in an optimum fashion has been approached in a variety of ways by researchers and practitioners. However, the Markov decision process has found very limited use despite the fact that cumulative damage is readily modeled by a Markov chain and that a wealth of immediately applicable theoretical results exist in the literature. The solutions are known for a variety of problems involving inspection, repair, and replacement, making it possible to solve directly for an optimal policy in the form of a control law. This paper reviews some of the relevant theoretical results in order to call them to the attention of civil engineers involved with pavement management systems.

A new F‐15 aircraft is now the heaviest fighter aircraft in the United States Air Force inventory. A mechanistic design procedure is developed for flexible pavements to accommodate the heavyweight F‐15 loading. The ILLI‐PAVE finite element computer program is used as the structural model to compute pavement structural responses. Design algorithms are developed from an ILLI‐PAVE data base and are demonstrated in an example problem. The proposed design procedure is then compared with the present U.S. Department of Defense (DOD) procedure. The DOD procedure does not adequately consider asphaltconcrete fatigue effects. DOD minimum surface thickness requirements may not be sufficient to prevent premature cracking of pavements subjected to long‐term use by the heavyweight F‐15 aircraft. However, the DOD procedure may be overly conservative for subgrade rutting.

Existing studies of transit network design deal separately with the problems of determining the optimal route spacing and operating headway, and that of determining the optimal stop spacing. This paper presents an analytic model for the design of an optimal feeder bus network for accessing an existing rail line, which avoids the sequential approach and combines these three basic variables. Our results with regard to bus‐route spacing and headway are similar to those obtained previously, indicating that route spacing and operating headway are not highly sensitive to changes in the relevant system parameters. With regard to bus‐stop spacing, three different cases are considered, reflecting three different stop‐spacing policies. In the first case, bus‐stop spacing is specified as uniform over the entire area. In the second case, stop spacing is constant along any given route. In the third case, stop spacing may vary both between and along routes. Closed‐form solutions are presented for the first two cases. The solution to the third case is shown to depend on an initial condition. Three numerical examples are presented, indicating that the proposed model provides reasonable solutions and consistent responses to changes in demand.

The current system of supply standards for buses in Korean cities fails to reach the demand for services. Passenger demand varies with route, time of the day, and with location along each route. In order to overcome this deficiency, alternative supply standards between the bus capacity and the passenger demand have been developed. The results of the analysis of this alternative reveal that some bus routes would gain and some would lose from the uniform application of the alternative supply standards. However, the overcrowding is spread out more uniformly through the alternative standards. Therefore, alternative standards may not be effectively applied under the present industry structure due to restrictions imposed by the uniform fare and to greater fluctuations in profitability.

This paper focuses on the effect on passenger dwell time from a complicated fare structure on a bus transit system. Analysis of data collected from three routes of the Southeastern Michigan Transportation Authority (SEMTA) is used as a case study. SEMTA uses a zone fare system. The nine different fare type and payment methods are found not to affect the dwell time significantly. Regression analyses are used to determine independent equations for the boarding and alighting times as a function of the number of passengers at a stop. These equations are compared to those found by past researchers. While the alighting times are similar, the boarding times are much less affected by the number of passengers at a stop. Using Kolmogorov‐Smirnov good‐ness‐of‐fit testing, the aggregate boarding and alighting times are also tested and found not to be significantly different from a lognormal distribution. The research concludes that some type of distance‐based fares could be used without significantly increasing the dwell times.