Story-by-story identification of multi-story frames from hysteretic response data

摘要

A parameter estimation technique for the identification of hysteretic behavior of lateral load resisting components of torsional shear buildings during seismic excitations is developed. The analysis of nonlinear structural response allows to assess the actual performance of lateral load resisting elements, such as energy dissipation characteristics and permanent deformations, during strong motion events. The proposed method aims to reconstruct the nonlinear force-displacement loops of individual frames by employing the measured acceleration of rigid floor and ceiling slabs as well as the knowledge of mass properties of the structure. The internal forces of the lateral load resisting elements are parameterized in terms of their drifts and the relative velocities between ceiling and floor slabs through an adoption of the well-known Bouc-Wen model. Such parameterization along with the equilibrium of the inertial and restoring forces acting on each floor are used to develop a state-space representation of the structure as a dynamic system whose state variables consist of the restoring forces as well as displacements of individual frames. An Unscentel Kalman filtering approach is subsequently implemented for the state-estimation of the dynamic system. The application of the proposed method is demonstrated through the analysis of simulated response of an idealized one-story shear building to 1989 Loma Prieta earthquake. The numerical verification studies indicate that the algorithm has promising potentials for characterizing various extents of nonlinearity experienced by individual frames in buildings with coupled lateral and torsional motion.