Subspace identification of Vincent Thomas Suspension Bridge ambient vibration data

In this paper the recently popular stochastic subspace identification (SSI) techniques are employed for the identification of modal parameters of the 766 m Vincent Thomas Suspension Bridge using ambient vibration data. The bridge, located in Los Angeles, is instrumented with a total of 26 fixed location strong motion transducers. This relatively small number of instruments introduces a low degree of spatial resolution in the analysis. As a result, mode shape information is extremely limited and difficulty is observed in identifying highly damped modes. Emphasis is placed on the extraction of damping and eigenfrequencies. The Canonical Variate Analysis (CVA) approach to covariance-driven stochastic realization is employed and an investigation of the modal parameters for the structure is made possible by examining results from two sensor groupings. Results indicate that for the data available the transverse modes contain the highest uncertainty whereas the vertical modes show consistency over varying size of the Hankel matrix.