Development of Linear Models to Predict Superstructure Ratings of Steel and Prestressed Concrete Bridges

Abstract

Due to the importance of estimating costs and approving budgets of maintenance and construction bridges, it is necessary to evaluate how some variables, such as structure material, average daily traffic, and length of spans, interacts among them, and how that relationship affects the costs. Therefore, this study is focused on the analysis of the longevity through the relationship among the different variables that the National Bridge Inventory (NBI) has in order to understand more of bridge superstructure rating. Particularly, this research is centered on what happens with Oklahoma's steel and pre-stressed concrete bridges and what variables are determinant of the bridge superstructure rating. To do that, by using the NBI database of 2010 and applying statistical methods such as Pearson's Correlation and Multiple Regression, linear models for predicting the superstructure rating of each material are developed. Although the bridge age and the average daily traffic have been the principal variables of deterioration in previous studies, the superstructure material has not been identified as an important determinant.