Model updating of nonlinear dynamic system using strong vibration data

• Main Author: Franciscus, Christiandy
• Other Authors: Cheung Sai Hung, Joseph
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/68842
• Institution: Nanyang Technological University

Reliability updating in nonlinear dynamic system is essential in structural health monitoring and structural control because uncertainties exist in the evaluation of structural dynamic parameter values. These uncertainties are caused by many factors, such as material variability, inaccurate construction process, material deformation due to earthquake and material deterioration throughout the structure lifetime. Using wrong parameter values in assessing the structural dynamic response could lead to invalid structural reliability. Together with nonlinear hysteretic response behavior of the structure, these uncertainties cause the nonlinear system reliability updating problem to be complex. Bayesian updating methods based on Bayes’ Theorem provide a way to model the uncertainties of the parameter values and have the ability to update them into a more reliable ones. It began to gain popularity and have been used in many fields, yet its application in structural models is very limited. This project aims to perform a recently developed Bayesian updating method, called Transitional Markov Chain Monte Carlo (TMCMC) method and discuss its effectiveness and efficiency in nonlinear dynamic system model updating. The method was applied to 5-story shear building subjected to twice 1940 El Centro earthquake, Southern California. 21 parameters, which included elastic stiffness, yield limit, post-yield-stiffness ratio, damping ratio of each story and measurement error, were updated from their prior beliefs into a more accurate ones, while nonlinear dynamic analysis of the structure was carried out by one of numerical time stepping methods, Newmark’s method. The application shows the ability of Transitional Markov Chain Monte Carlo to carry out model updating in nonlinear dynamic system.