Effect of sensor system noise and load positioning on the precision of load testing and rating of highway bridges: a case study

This paper discusses the load testing of a three-span simply supported multi-girder/stringer bridge superstructure in northern New York. The first part of the testing deployed a custom developed wireless sensor system (WSS) measuring high-rate real-time strain measurements on all three spans of the steel superstructure. Controlled progressive damage was introduced to one of the three spans for the second part of testing where the span was monitored using wired strain transducers mounted at the top and bottom flanges of each girder at midspan. A single test vehicle was used to apply prescribed loads to the structure to acquire estimates of neutral axis (NA) and transverse load distribution factors (DF), two common parameters typically measured for a load rating. A sensitivity analysis was conducted using the peak-to-peak noise in the measurement signals. As expected, the farther the axle loads were positioned from the measured girder, the lower the measurement signal. In certain cases, this produced significant uncertainty in the NA measurement. On the other hand, the peak-to-peak noise did not appear to affect the DF estimates. This data was combined to illustrate the variation in the overall inventory and allowable stress design (ASD) load rating of the steel superstructure.