Dynamic model updating with applications in structural and damping systems: From linear to nonlinear, from off-line to real-time

Wei Song, Purdue University

Abstract

Dynamic model updating is a technique to investigate the status of a structure by using dynamic information identified from structure vibrational data. By satisfying certain cri-teria based on modal information of the structure, the dynamic properties of a mathemati-cal model can be updated to match those from the measurements of a physical specimen. Based on the purposes of model updating, the updating criteria, the complexity of the structural models to be updated, and the types of dynamic information sought, model up-dating methods can be categorized into the following: i) linear finite element (FE) model updating using modal information, ii) nonlinear FE model updating using modal informa-tion based on a linearized model under certain operation conditions, zero loading point in this study; and iii) nonlinear hysteretic model using time-series data for real-time updating. The focus of this dissertation is to develop the above three different updating methods and demonstrate their effectiveness in corresponding updating applications. Both linear and nonlinear hysteretic models are considered in the study. Numerical simulations are carried out for all the three model updating techniques to demonstrate the performance and efficacy of the updating techniques developed. Two sets of experiments are con-ducted, one set contains real-time updating tests on magnetorheological (MR) dampers and another set contains quasistatic cyclic tests and real-time updating tests using shake table on a steel shear building. The real-time feature is achieved by the latest high per-formance host-target environment enabled by xPC Target™, a product from MATH-WORKS®. In addition to the development of the model updating techniques, a power supply unit model and a nonlinear hysteresis model modified from Bouc-Wen model are proposed and successfully applied in the experimental study. The study conducted herein demonstrates that the model updating techniques developed can be effectively applied to various updating scenarios. The developed nonlinear hysteretic model updating technique is able to achieve “hard” real-time functionality, which can provide the most up-to-date information of the structural model. This technique also requires minimal data-buffer, and has potential to have a significant impact on nonlinear structural control applications.

Degree

Ph.D.

Advisors

Dyke, Purdue University.

Subject Area

Civil engineering|Mechanical engineering

COinS