Date of Award

12-2017

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Civil Engineering

Committee Member

Dr. Sez Atamturktur, Committee Chair

Committee Member

Dr. Hsein Juang

Committee Member

Dr. Weichiang Pang

Committee Member

Dr. Nadarajah Ravichandran

Committee Member

Dr. Akshay Gupte

Abstract

Early detection of the deterioration and degradation in civil infrastructure is critical for structural engineers and infrastructure managers to develop rehabilitation and maintenance plans. In the field of structural health monitoring, numerous techniques have been developed to detect and localize damage by examining changes in measured vibration response. Among vibration-based damage detection techniques, model-based approach has been widely used as its damage detection process incorporates the geometric configuration, physical properties, and behavioral characteristics of the structural system. However, the model-based approaches depend on a model calibration procedure that is based only on the outputs of numerical models without explicitly taking the knowledge regarding the mechanistic behavior of the system into account. Moreover, due to the limitation of measurement degrees of freedom (DOFs), the number of identified vibration modes are typically far fewer than the number of model variables to be calibrated. Consequently, these model-based damage detection methods frequently suffer from an ill-posed inverse-problem. This dissertation contributes to the field of model-based damage detection by implementing the Extended Constitutive Relation Error (ECRE), a method developed for error localization in finite element models for detecting structural damage. Implementing ECRE for damage detection leads to the localization of elements with high residual energy through the identification of discrepancies between experimental measurements and model predictions due to damage. The ECRE-based damage detection technique incorporates the underlying physics of the problem in a tangible and visible manner, and thus leading to more reliable solutions in the damage detection and localization process. This dissertation applies the ECRE-based damage detection in the context of both linear and nonlinear dynamical systems. In particular, the dissertation integrates the Multi-harmonic balance method with ECRE to accurate identify the modeling errors of locally nonlinear dynamical systems. This approach has a potential to be applied for damage detection in the nonlinear structural system, as well as to be used as a damage prognosis tool for the estimation of structural system's remaining useful life.

Share

COinS