Structural health monitoring: Damage detection and localization using changes in stiffness properties

The objective of this project was to develop a damage detection technique that can be applied to structural health monitoring of structures. The researchers at Utah State University have performed full-scale vibration testing on various civil structures in recent years. This research project has shown the feasibility of determining the integrity of testing structures by examining the changes in dynamic and stiffness properties. Recently, increased interest has been focused on the utilization of vibrational techniques in the fields of nondestructive evaluation and damage detection. Damage that has occurred in a structure will change the modal frequencies and modal shapes. A methodology to localize damage in structures by changing the stiffness properties was studied by using a finite element model.; We proposed an algorithm to determine the eigenvalues and eigenvectors matrix from the testing vibration data using the Eigensystem Realization Algorithm with an Observer Kalman Filter Identification (ERA-OKID). The identified frequencies obtained from ERA-OKID are used as the guidelines for the finite element model. The measured modes (eigenvalues and eigenvectors) use optimal correction of stiffness matrices. This approach assumes a linear structure both before and after damage, with damage being considered as a single point in time.; This dissertation is roughly divided into three parts: Chapter 2 provides typical background on dynamic response, modal update process, and structural health monitoring; Chapters 3 to 5 deal with a methodology to the problem of dynamic analysis and to damage detection; Chapter 6 reports selected case studies on demonstrating the success of the proposed damage detection algorithm and localization approach.