Reliability updating of nonlinear dynamic system using strong vibration data

• Main Author: Tee, Lay Sin
• Other Authors: Cheung Sai Hung
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Civil engineering
• Online Access: http://hdl.handle.net/10356/72976
• Institution: Nanyang Technological University

Reliability updating has become an essential tool to predict structural responses under dynamic loads. However, there are many uncertainties in structural properties and measurement error exists in the model. For this reason, the outcomes of the experimental and theoretical model are different. Hence, the analytical models of the structures need to be updated based on the experimental results. In general, majority of the method uses model analysis to get unknown parameters. There are many methods available for the reliability updating. In this study, Bayesian theorem and Subset Simulation method are combined to quantify the uncertainties and errors in the model parameters. Here, Bayesian probabilistic methodology is integrated with probabilistic analysis tools for the purpose of updating the assessment of the robust reliability based on dynamic test data. However, evaluating the integral equation may be difficult if the product of the likelihood and prior is complex. For instance, the parameter space is high dimensional which requires computing high dimensional likelihoods. In order to solve the complicated computational problems, method for updating the structural reliability is proposed. The Subset Simulation method is adopted to solve the problem of dynamics system involving high dimensional parameters. In this project, the proposed Subset Simulation method will discuss the effectiveness in nonlinear dynamic system model updating. The algorithm will be applied upon a 5-storey shear building model. This model was tested based on 21 unknown parameters, which include elastic stiffness, yield limit, post-yield-stiffness ratio, damping ratio of each story and measurement error. It is observed that the Bayesian theorem with Subset Simulation technique will give a better estimate of unknown structural parameters that coincide to the actual values. Finally, the Subset Simulation method was performed to update 21 unknown parameters successfully.