Structural damage detection using Empirical Mode Decomposition and HHT

The importance of developing robust systems that can detect and locate progressive deterioration in structures or abrupt damage induced by extreme loading events is well recognized in the field of structural health monitoring. A reliable damage detection algorithm is the critical element of this development. This paper discusses a new time-frequency data analysis method, Empirical Mode Decomposition (EMD) and Hilbert-Huang Transformation (HHT), and its application to damage detection. The time series data from numerical simulation and laboratory experiments of a simple structure with and without damage are processed to determine the presence and location of structural damage. This is done by extracting a set of basis components directly from the measured response of a system and tracking phase properties between successive degrees of freedom of the structure. Our results illustrate that this new approach, along with simple physics-based models, permits the development of a reliable damage detection methodology.