Launching Semi-automated Modal Identification of the Port Mann Bridge

Abstract

Southwest British Columbia (B.C.) is home to many of the largest bridges in North America, and the majority of its lifeline bridges are cable-stayed. It is also one of the most active seismic zones in Canada and the world. To help face the obvious challenges that this poses for structural engineers, the Ministry of Transportation embarked on a province-wide monitoring program in 2009, called the B.C. Smart Infrastructure Monitoring System (BCSIMS). It combines the vibration data from an urban strong motion network and a structural health monitoring network, and pursues two objectives: (a) to enable an efficient emergency response by implementing a post-earthquake damage assessment module, and (b) to install a long-term monitoring module for cost-effective operation over the entire lifespan of structures. This paper contributes to the post-earthquake damage assessment module and focuses on the modal identification of the Port Mann Bridge using the stochastic subspace identification (SSI) method. Moreover, a hierarchical clustering approach is used to automatically select stable modes of vibration. The envisaged damage assessment module is based on subspace-based methods, and the modal parameter estimation is merely a side product of the ongoing studies. Nonetheless, it is an integral part of every structural health monitoring network and allows first insights into the influence of environmental/operational variables on the dynamic behaviour of the bridge when, for example, combined with regression analysis. No bridge weigh-in-motion system is installed, but a strong correlation can be found between a traffic index estimated from toll cameras and the fluctuation of modal frequencies. This highlights the dominant effect of traffic loads over other environmental variables.