Nonlinear Finite Element Model Updating of a Large-Scale Infilled Frame Structures Based on Instantaneous Modal Parameters

Abstract

While linear finite element (FE) model updating has been successfully applied for predicting structural damage as loss of effective stiffness, nonlinear FE model updating can provide improved and more accurate damage identification (i.e., a more comprehensive measure of damage) and can additionally be used as a tool for damage prognosis. The current study is focused on characterizing the nonlinear material behavior in a FE model of a three-story infilled frame using the identified instantaneous natural frequencies and mode shapes. The 2/3-scale, 3-story, 2-bay reinforced concrete frame with masonry infills was subjected to large amplitude earthquake base excitations on a shake table. The deterministic stochastic subspace identification method is used for estimating the instantaneous (during short-time windows) modal parameters of the structure based on the nonlinear response of structure during a seismic base excitation. Parameters of a priori selected hysteretic models (Bouc-Wen) at different finite elements of the structural model are calibrated to minimize the misfit between the identified modal parameters and those from the FE model. The accuracy of the calibrated FE model is assessed through the comparison of the predicted response and natural frequencies obtained from the model with those of the specimen.