This paper presents two novel damage indices based on empirical mode decomposition
(EMD) and fast Fourier integration for identifying structural damage caused by a
change in structural stiffness. The paper also demonstrates the effectiveness
of the proposed damage indices formulated based on a series of coupled
mathematical/engineering approaches that are used to detect damage in pipes
reliably and accurately. The main approach is based on monitoring the vibration
response of pipes using piezoelectric sensors and the first intrinsic mode functions
(IMFs). Finite element analysis is used to simulate the response of a healthy pipe, as
well as pipes with various sizes of damage. Damages are meant to represent the
outcome of local corrosion (damage) with varying reduction in areas around the
circumference of the pipe. The evaluated damage indices could effectively establish
the location of the defects. Moreover, the evaluated energy indices could also
distinguish various size defects. To demonstrate further the effectiveness of
our proposed damage indices, the results are compared with other effective
indices based on wavelet packet and other statistical methods reported in the
literature.
Keywords
damage detection, empirical mode decomposition, wavelet,
vibration based, health monitoring
