Electronic Theses and Dissertations

Identifier

6025

Date

2017

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Engineering

Concentration

Civil Engineering

Committee Member

Charles V Camp

Committee Member

Shahram Pezeshk

Committee Member

Adel Abdelnaby

Abstract

A new method is developed for finite element model calibration of structures with the results of modal testing. The proposed method applies multi-objective optimization to develop a set of calibrated models and employs sensitivity analysis to analyze and identify the most effective parameters for model calibration. The study consists of a full experimental study on modal identification of structures under ambient vibration conditions and an analytical study on finite element model calibration. The experimental study is focused on operational modal analysis of structures with covariance driven stochastic subspace identification in the time domain and frequency domain decomposition in the frequency domain. In the analytical part, the model calibration problem is defined as an optimization problem with the objective of minimizing the discrepancies between the modal frequencies and mode shapes identified from the test and estimated from the finite element model of the structure. Single objective and multi-objective formulations are developed for the model calibration problems and evolutionary algorithms are applied to solve these optimization problems. In order to reduce the complexity of the optimization problems and improve the quality of the calibration results, a variance-based sensitivity analysis is applied to examine the effectiveness of the updating parameters and remove unnecessary parameters from the calibration process.The effectiveness of the applied methodology is examined by conducting experimental and analytical studies on a three-span highway bridge on the Interstate 385 in Arlington, TN. Following the experimental study, a detailed finite element model of the bridge is developed and prepared for sensitivity analysis and model calibration with the proposed optimization techniques. A comparison of results is presented between the model calibration from single objective and multi-objective optimization. In addition, model calibration results with and without sensitivity analysis are presented to examine the effectiveness of this method. Finally, a comparison is presented of model calibrations for different cases.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to the local University of Memphis Electronic Theses & dissertation (ETD) Repository.

Download

Share

COinS