Journal of Infrastructure Systems

A basis for the direct use of data from nondestructive evaluation methods in bridge management systems is presented. Bridge management systems use integer-valued condition ratings to recognize conditions of bridge elements, to model progression of deterioration in elements and to determine needs for maintenance and repair. In an ongoing research project funded by the U.S. Federal Highway Administration, procedures are being developed to communicate data from quantitative nondestructive evaluation methods to bridge management systems. Data from nondestructive evaluation methods can inform management systems on the extent of damage, on the initiation of deterioration processes and on the exposure of bridge elements to aggressive agents. In addition, data obtained through nondestructive evaluation allow the formation of models of specific deterioration processes in bridges. The use of these data in bridge management systems requires redefinition of condition ratings together with the creation of procedures for automated interpretation of the data. By these actions, nondestructive evaluation methods are used directly to assign condition ratings, and condition ratings are made into a terse report of nondestructive evaluation data that are compatible with management systems. This paper reports work in progress to integrate nondestructive evaluation methods and bridge management systems.

A method for estimating transition probability is applied to the deterioration models of a bridge management system. Deterioration is modeled as a time-dependent function of the condition changes of critical elements in bridge systems. The methodology applies “life data” analysis techniques employed in reliability studies of engineering systems and incorporates the inherent censorship of the utilized bridge data. It involves a two-phase process: one concerned with the condition of a set of similar bridge components composing a system, and the other concerned with the condition of a specific component within the system. The information required for modeling purposes is derived from historical data typically available in bridge inventory and inspection systems. The methodology focuses on deterioration models previously created for use in a bridge management system. It is illustrated by a case study involving deterioration processes of a structural deck system. The methodology has been evaluated through comparison of predicted and realized condition measures over a 10-year period. On the basis of the analysis conducted and its results, it is concluded that the developed methodology significantly improves the accuracy of bridge deterioration models and the network-level condition prediction process.

A universal-type model, taking into account the various possible forms of deterioration rates during a pavement's life span, was introduced by the writer elsewhere. The subsequent deterioration curves for each cycle of the overlay placement were derived from the parameters of the original deterioration curve for any specific maintenance and rehabilitation (M&R) strategy applied to the pavement under consideration. The various computations involved with these deterioration curves, however, were found to require a considerable amount of computer time and memory space. Therefore, shortcuts for these calculations were developed, consisting mainly of correlative equations that can predict in a straightforward manner (1) the age of the pavement for any given condition; and (2) directly from the previous cycle, the subsequent parameters of the deterioration curve for any overlaid cycle.

The relationships between axle weight and pavement damage are traditionally based on data collected from road tests. The goal of the research presented in this paper is to use available field data to develop nonlinear dynamic simulation techniques for predicting pavement performance. The study provides information on pertinent issues through models of vehicle dynamic loads and their effect on selected pavement structures. The model structure allows estimates of the cracking and failure of rigid pavement slabs and joints as a result of repeated applications of the dynamic loads simulated for specific gross vehicle loads and tractor-trailer configurations. A preliminary pavement consumption model is proposed that gives reduced equivalent pavement thickness as a function of number of vehicle passes.

The State of Illinois has been conducting Condition Rating Surveys (CRS) for nearly 30 years. The results of these condition surveys have become the primary vehicle for providing the executive and legislative offices with an assessment of the condition of the state's pavement network. The CRS has also become the cornerstone of the Illinois Department of Transportation (IDOT) pavement management system. The results are used extensively at the network level to determine budget share for the pavement improvement program and at the project level to determine priorities of specific pavement improvement proposals. The CRS initially involved a panel of four to five technical persons driving over each pavement section in the network to subjectively assign a numerical value ranging from 1.0 for a pavement in failed condition to 9.0 for a pavement in excellent condition. In 1989, several concerns prompted IDOT to investigate the feasibility of automating the collection of pavement condition data. These concerns included the safety of the expert panel; the reduction in staff and hiring restrictions being placed on the agency; and the total subjectivity of the existing assessment methodology. In 1994, IDOT adopted new automated technology to conduct the surveys. At the same time, IDOT hired a consultant to develop equations that could be used with identified predominant distresses to calculate a CRS value that falls within one-half point of the rating that would have been obtained through the manual on-site surveys. This paper documents the successful adoption of new technology, development of CRS models, and integration of the CRS models into the department's program development process. The changes brought about by automating the CRS process also increased the importance of properly identifying distress types and extent for each pavement section. As a result of this change, the department identified the need for additional training that could be used to supplement the annual training class and could provide immediate support to the CRS raters as they performed the condition assessment surveys. The paper also documents the development of the CD-based interactive training program that was developed for the department. The training program features an overview of the CRS process and individual distress identification.

A joint effort between the Connecticut Department of Transportation and the University of Connecticut has been underway since 1984 to use different monitoring approaches in the assessment of the state's bridge infrastructure. This program has involved both short-term and long-term studies to provide information on performance under routine traffic of a wide spectrum of bridge types. Most short-term studies have been used to provide information on connections, members, and diaphragms. These were done to assess the need for replacement or repair or to ascertain that the component or portion of the bridge was behaving as designed. The long-term studies have been conducted to develop information on global behavior, both for evaluation purposes and to study approaches which may be used for identifying structural integrity problems. This paper reports on the history of this bridge monitoring program, with the major findings. It also reports on the beginning of a multiphase effort to install continuous monitoring systems on different types of bridges in the State of Connecticut. The goals are to demonstrate the feasibility of using monitoring systems to supplement field evaluations and to collect information that will be used in the management of the state's in-service highway bridges.