Probabilistic structural integrity evaluation of a highway steel bridge under unknown trucks

Abstract This paper presents fragility curves derived from a best-fit regression model that enable to quantify probabilistic structural integrity of an in-service highway steel bridge under multiple truck passages with uncertain characteristics. The regression model is able to be identified via analytical modelling techniques incorporating the bridge response quantities resulting from unknown five-axle trucks through structural health monitoring system. These quantities coupled with weigh-in-motion (WIM) data obtained from two weigh stations closest to the bridge are used (1) to identify which unknown trucks are presumably to travel over the bridge and (2) to quantify the plausible characteristics of the trucks and the corresponding load ratings determined following the AASHTO Manual. With this information, the regression models can be made to demonstrate load ratings as the truck characteristics change. Based on the best-fitted load ratings, nine significant truck characteristics were looked at in their role in causing a below satisfactory bridge capacity. Key findings reveal that the all axle weights affect the fragility curves though the axle weights are not significantly different, and the most significant axle spacing is the first spacing.