A decision support methodology for the rehabilitation and replacement of highway bridges

Abstract

The current rate of deterioration of highway bridges in the United States has created the need for a comprehensive bridge management system (BMS). This need has also been recognized in Texas, necessitating the development of a comprehensive bridge management system (BMS) in Texas, transportation agency. This dissertation presents the results of a study conducted to develop analytical tools required as part of a BMS. First, a preliminary study was conducted to model the deterioration patterns of Texas highway bridges--bridges under the jurisdiction of the Texas State Department of Highways and Public Transportation (SDHPT). The second part of the dissertation discusses the development of a computerized decision support methodology for the evaluation and selection of feasible improvement strategies--maintenance, rehabilitation, and replacement - on a bridge. Using the bridge condition data available for Texas bridge on the SDHPT's Bridge Inventory Inspection and Appraisal Program (BRINSAP) computerized database and the Federal Highway Administration (FHWA)'s records, statistical correlation analyses, regression analyses, and probability models include linear, piecewise-linear, and nonlinear models while the probability analyses resulted in state probability and age probability distribution. In addition, the expert opinions of Texas bridge engineers were elicited and used to formulate linear models of bridge deterioration. Once a bridge deterioration pattern is modeled, it is the possible to predict furfure condition and deficiencies on the bridge. Feasible strategies should be synthesized to correct of at least halt the observed or predicted deterioration of the bridge. To aid the bridge engineer in evaluating and selecting the best of these feasible improvement strategies, a decision support methodology based on a systems approach was formulated, capable of incorporating the expert judgment of bridge engineers...