Fatigue is one of the most critical problems for steel bridges as well as for any steel structures that needs to be considered during design and operation. The objectives of this study are to explore monitoring technologies, and to develop effective structural and data analysis strategies as well as implementation recommendations for evaluating performance of fatigue-sensitive details and retrofits in steel bridges. Acoustic emission (AE) was selected as a candidate inspection technology, and a monitoring system was installed on a bridge. In general, the performance of the monitoring system and associated software is satisfactory. The majority of AE monitoring challenges are associated with AE data analysis and interpretation of results. In this study, cluster analysis and non-linear mapping signal analysis techniques are used to group AE data with similar waveform characteristics. The presence of the signals that resemble the characteristics of crack opening signals, noise, and structural resonance is identified through waveform analysis. Once the presence of crack opening signals is confirmed, the source location plots are utilized to assess the concentration and the level of activity at the locations of interest.

A significant difference is observed in the fatigue life calculated using measured stress, and the stresses calculated using finite element models loaded with a fatigue truck. Hence, a two-tier implementation process is recommended. Tier I process includes the assessment of bridges with repaired details. Additional recommendations include implementing AE data interpretation capability in an on-line system to provide reliable input with minimal interpretation requirements for inspection-based maintenance management.