Parametric Estimation of Wave Dispersion for System Identification of Building Structures
The linear-elastic response of a building structure subjected to an earthquake base excitation can be approximated as the response of a continuous, spatially inhomogenous, dispersive, viscoelastic solid subjected to vertically incident plane shear waves. The frequency-dependent phase velocity and attenuation of seismic energy at different wavelengths, together with the inertial properties of the multilayer solid characterize the response of the building structure. The objective of this study is to identify the structural system by estimating the parameters that characterize the propagation of seismic waves in an equivalent multilayer viscoelastic solid. To pursue this objective, first, the measured dynamic responses of a building structure are used to derive the frequency response functions (FRFs) of the floor absolute acceleration with respect to the base excitation using a seismic interferometry approach. The FRFs obtained from the measured structural responses are then compared with the FRFs estimated using analytical models for one-dimensional shear wave propagation in a multilayer Kelvin-Voigt dispersive medium. Through a recursive Bayesian estimation approach, the parameters characterizing the phase velocity and damping ratio of the multilayer medium are estimated. This study provides a step forward in seismic interferometric identification of building structures by proposing a new method for parametric estimation of shear wave velocity and damping dispersion at the story level of a building structure. The estimated shear wave velocities before and after a damage-inducing event can be used to identify permanent loss of effective lateral stiffness of the building structure at the story level, thus can provide an alternative method for structural health monitoring and damage identification.
Additional Information© 2017 Springer International Publishing AG. Support of this research provided by the Terrestrial Hazard Observation and Reporting (THOR) Center at Caltech is gratefully acknowledged.
Accepted Version - 166_Ebr_Intended_to_Read.pdf