Correlation between reduction in load capacity and structuralcondition of highway bridges

摘要

Although it is generally acknowledged that the load capacity of a bridge tends torndecrease as a bridge deteriorates, there exists no consensus about the appropriate quantitativernform of such a dependency. Consequently, existing bridge management systems, such as Pontis,rntypically do not model a reduction in load carrying capacity as a function of age or deteriorationrnof bridge elements. An analysis of the 2003 U.S. National Bridge Inventory (NBI) data hasrnbeen performed to evaluate the statistical relationship between load capacity (measured usingrnthe inventory rating) and five variables: age of the bridge (measured in years since the bridgernwas constructed or reconstructed); design load; and NBI condition ratings for the bridge deck,rnsuperstructure and substructure. The regression analysis was conducted on set of over 468,000rnbrides, including all bridges in the NBI with complete data except for culverts, pedestrianrnbridges and railroad bridges. The analysis indicates the most strongest predictor of load carryingrncapacity is the design load of the bridge. The correlation coefficient r is 0.55 or greater betweenrndesign load and load capacity. The correlation coefficient r is between 0.27 and 0.40 betweenrnload capacity and age and the condition ratings. Design load is highly correlated with agernand the condition ratings, implying that bridges with higher design loads tend to be newerrnbridges in better condition than bridges with lower design loads. When design load is factoredrnout as a predictive variable, the resulting correlation observed between load capacity and thernremaining variables is weak, with a correlation coefficient r of 0.2 or less. The analysis supportsrnthe conclusion that there is no clear quantitative relationship in the available NBI data betweenrnload capacity and age of a bridge, or between load capacity and physical condition.