Operational Modal Analysis of a Footbridge by Fast Bayesian FFT Method

This paper presents work on field instrumentation and modal identification of Footbridge NF276 (Highways Department, HKSAR) spanning across the Tolo Harbor in the Eastern New Territories, Hong Kong. An array of tri-axial acceleration measurements under ambient conditions was deployed. A total of 74 locations were measured tri-axially, giving 74 x 3 = 222 degrees of freedom (dofs). A number of setups were carried out to cover all locations of interest using a limited number of sensors with all setups sharing a common set of reference channels. In every setup, five minutes' data was collected and used for modal identification. A Fast Bayesian FFT method is used for modal identification using the measured ambient data. The approach views modal identification as an inference problem where probability is used as a measure for the relative plausibility of outcomes given a model of the system and measured data. The mode shape of the whole bridge is obtained by assembling those identified from individual setups using a least square approach. The identification results reveal a number of interesting features about the structural system. The (frequentist) setup-to-setup variability of the identified modal parameters shall be investigated and their practical implication shall be discussed.