Bridge rating using in-service data in the presence of strength deterioration and correlation in load processes

Abstract

This paper presents a probability-based methodology for load rating bridges that can accommodate detailed site-specific in-service structural deterioration and response data in a load and resistance factor rating (LRFR) format. The use of site-specific structural response allows the elimination of a substantial portion of modelling uncertainty in live load characterization. Inclusion of structural ageing allows the bridge owner the choice to rate for longer intervals than, say, the usual two-year inspection cycle. This methodology allows the live load-effect sequence on bridges to be statistically stationary with a weakened mixing-type dependence that asymptotically decreases to zero with increasing separation in time, instead of making the common assumption of independent and identically distributed sequences of live loads. In addition, uncertainties in field measurement, modelling uncertainties and Bayesian updating of the empirical distribution function are considered to obtain an extreme value distribution of the time-dependent maximum live load. Gross section loss due to corrosion occurring with a random rate governed by an exponentiated Ornstein-Uhlenbeck type stochastic noise is considered. An illustrative example utilizes in-service peak strain data from ambient traffic collected on a high-volume steel girder bridge. In-service load and ageing resistance factor rating (ISLARFR) equations corresponding to plastic collapse of critical girder cross-section over a range of service lives are developed.