Singular spectrum analysis based structural damage identification in beams with multiple breathing cracks

Abstract

This paper presents a generalized vibration-based multiple breathing crack localization technique in beams with an unknown number of cracks based on Singular Spectrum Analysis (SSA). Localization of multiple breathing cracks is a highly challenging inverse problem, as all these breathing cracks (more than two) might be in a similar state (either opening or closing state) at any particular time instant or in contrasting states (while some breathing cracks are opening, others in closing state) during vibration. The level of nonlinearity of vibration response is strongly dependent not only on each crack size, and location but also on the application of driving force along the cracked beam. The concept of varied input frequency excitation sources is employed in the present work for multiple breathing crack identification over the tedious traditional approach of varying input force application positions. The major advantage of using singular spectrum analysis in the present work is that it can reliably isolate and extract the very low-amplitude nonlinear sensitive components (super harmonics and intermodulation) being buried in the total response based on the pairwise eigenvalue property of harmonic components. Investigations have been carried out by varying the number, spatial locations, and also intensities of the breathing cracks. Sensitivities associated with measurement noise and also with limited sensors are also investigated. The results of both numerical and experimental investigations carried out in this paper concluded that the proposed SSA can effectively localize closely spaced or sparsely spaced (i.e.,, spatially far apart) multiple cracks in the beam. The proposed SSA approach is data-driven, works with limited sensors, and does not need reference healthy state measurements.

Keywords:

• breathing crack
• harmonic vibration
• multiple cracks
• super harmonics
• singular spectrum analysis

Acknowledgements

The author thanks the technical personnel at CSIR SERC's SHML and ASTAR Lab for their assistance and support in conducting the experiment.

Disclosure statement

No potential conflict of interest was reported by the authors.