Development of performance assessment tools for a highway bridge resulting from controlled progressive monitoring

As highway bridges continue to deteriorate given the increased service life, increase in vehicle demand and exposure to harsh environmental climates, new methods of monitoring their in situ performance are of high priority. Damage within the structure can alter various load demand and capacity characteristics, affecting the overall integrity of the bridge. Discussed in this paper is the monitoring of a simple span bridge superstructure under various induced damage states. Strain measurements were recorded at the midspan and north abutment of each girder. Six levels of damage progression were implemented at a rocker bearing and various diaphragms to girder connections. Transverse load distribution factors (DFs) and neutral axis (NA) locations were measured for each damage case and evaluated against the baseline undamaged response. These measurements serve to provide a possible method of damage detection using load-testing parameters already employed by various transportation agencies. Next, a performance index (PI) is developed for this stringer/multi-girder bridge utilising the NA and DF response from the steel girder system and the allowable stress design load-rating data. The ratio of NA to DF was compared to the inventory load rating for each girder at each damaged state. The data were fitted with a power regression model to form the PI. Furthermore, a 95% prediction interval was used around the predicted response to capture all the data from the testing. The model was applied to the damaged structure as well as two additional stringer/multi-girder bridges. The objective of the PI is to complement existing qualitative assessment protocols with quantitative results for improving the condition assessment process.