Z.W. Chen, et al.
Bridge Maintenance, Safety, Management, Resilience and Sustainability – Biondini & Frangopol (Eds). London: Taylor & Francis Group, 2012: 737- 744.
Many long-span cable-supported bridges have been built throughout the world in recent years. They begin to deteriorate once built and continuously accumulate damage during their service life due to natural hazard and harsh environment such as typhoons, earthquakes, vehicles, temperature and corrosion. Meanwhile, structural health monitoring technology gains a rapid development recently. Comprehensive structural health monitoring systems (SHMS) have been designed and installed in a number of long-span bridges, and different types of sensors are used for monitoring the loading, responses and conditions of bridges. A well-known example is the Wind and Structural Health Monitoring System (WASHMS) installed in the Tsing Ma suspension bridge in 1997. It is a new trend to integrate SHMS and damage detection technology for real-time condition assessment of long-span bridges. On the other hand, bridges failure often begins with minor local damage in bridge components. Therefore, an efficient and effective damage detection method sensitive to local damage is essential for bridge health monitoring. It has been recognized that conventional vibration-based damage detection techniques, commonly based on modal properties or their derivatives, are often insensitive to structural local damage and are significantly dependent on the change of operational environment, such as temperature. This paper explores a novel damage detection technique based on stress influence lines in representative bridge components, and its efficacy is validated through a case study of Tsing Ma Suspension Bridge. A mathematical method with regularization is first introduced to identify the stress influence line (SIL) based on the in-situ measurement of train information and train-induced stress responses in local bridge components. A good agreement between the identified and baseline SIL demonstrates the effectiveness of the proposed identification method. A damage index based on SIL is subsequently proposed and applied to a hypothetic damage scenario in which a typical local component is damaged. Train-induced dynamic stress response in the damage scenario is simulated numerically using coupled train-bridge dynamic method. The SIL in the damage scenario is identified from the simulated responses. The damage index is computed based on the SILs in the undamaged and damaged cases. The results indicate that the proposed method offers a promising technique for real-time damage detection of long-span cable-supported bridges equipped with comprehensive SHMS.