Wavelet transform-fractal dimension based methodology for structural damage localization

Abstract

This paper proposes a reference-free vibration-based structural damage localization technique using re-deployable mobile sensor concept and statistical signal processing algorithms. The re-deployable mobile sensor concept computes the mode shapes with the high spatial resolution by using only two sensors in contrast to the traditional high fidelity static dense sensor network which incurs huge costs and is resource-intensive. The two sensors are moved independent of the input traffic loading (position, speed and time of the vehicle covering the entire span), by progressive redeployment in various segments along the monitored structure. The local acceleration measurements are then postprocessed by a decentralized modal analysis technique to obtain a continuous global mode shape. For Structural damage localization using mode shapes, this paper employs a hybrid Continuous Wavelet Transform (CWT) – Fractal Dimension based methodology using only the output response measurements of the structure with damage. The data fusion of Katz Fractal Dimension (KFD) with CWT mode shapes help in suppressing the noise, representing the mode shapes in an enhanced time-frequency domain and intensifies the local singularities caused by the damage. Numerical and experimental investigation carried out on simple beam-like structures demonstrate that the proposed damage localization index based on CWT-KFD can able to localize and relatively quantify the damages including minor cracks and damages at more than one spatial location.