The detection of damage in structures through the use of vibrational methods offers particular advantages, which makes it an attractive method to use in specific applications. In this work the advantages and some of the possible applications of vibrational damage detection methods will be discussed. A study of the field of damage detection using vibration techniques is undertaken. Available methods are categorised in general groups according to the underlying principles. The principle, on which each group functions, as well as the advantages and disadvantages of each, concerning the practical application thereof, is explained. The goal of this work, which entails developing a damage detection method using large amounts of raw data directly and combining some of the most favourable properties of the different groups to detect damage, are set. The new method is developed and compared both numerically as well as experimentally to two methods, chosen from the literature because of similar methodology and their reputation for effective damage detection. The methods will be tested numerically with respect to accuracy, sensitivity and multiple damage detection ability. Finally experimental data is gathered and used to verify the methods damage detection ability. The new method provides a different approach to damage detection, by combining an available vibration detection method with the maximum available amount of data in order to increase the damage detection ability.
Dissertation (MEng (Mechanical Engineering))--University of Pretoria, 2006.