Detecting damage in beams and structures through modal analysis

• Main Author: Downer, Lee Matthew, 1983-
• Other Authors: Memorial University of Newfoundland. Faculty of Engineering and Applied Science
• Format: Thesis
• Language: English
• Published: 2010
• Subjects: Modal analysis; Structural analysis (Engineering); Poles (Engineering) > Defects; Damping (Mechanics); Newfoundland studies; University of Newfoundland
• Online Access: http://collections.mun.ca/cdm/ref/collection/theses5/id/17132

Thesis (M.Eng.)--Memorial University of Newfoundland, 2011. Engineering and Applied Science Bibliography: leaves 203-206. Based on a review of previous literature on the subject of modal testing, it was determined that modal parameters such as frequency, damping ratio and mode shape change with the introduction of damage to a beam or structure. However, relating those changes back to the exact nature and location of the damage is a subject of ongoing study. In the current work, a method has been proposed for quantifying and localizing defects in structures using multiple regression models and response surfaces obtained through design of experiments (DOE) techniques, which are initially developed to relate modal frequencies to parameters such as defect location and defect depth. Once the models are developed, multiple models can subsequently be inverted and solved for the multiple defect parameters required to characterize a defect by using modal frequency measurements of a test specimen. The method was also successfully employed in many scenarios involving theoretical, finite element and physical models. In addition to the development of this method, a series of full-scale utility poles were tested in order to investigate whether modal impact testing could be used to assess their condition. Static destructive tests were used to determine material properties as well as failure stress at the ground line and break location for each pole. It was found that each modal damping ratio correlated to some degree with these maximum stress values. Moreover, it was found that the average of damping ratio across multiple modes correlated with maximum stress better than either individual damping ratio, and that correlation progressively improved as a greater number of modes were considered in the averaging process. Regression models were developed to relate average damping ratio to maximum stress and proved to provide better predictions of maximum stress for the specimens involved in the study than did commercial ultrasonic NDT equipment.