Wavelet Based Damage Identification Of Structures

Wavelet transform and multi-resolution analysis are relatively new mathematical methods which can be well used in the research field of damage identification, while which is the kernel for structural health monitoring.Continuous wavelet transform (CWT) can be used to detect exactly the location of open cracks in a damaged beam. And meanwhile continuous wavelet transform can be used to estimate the Lipschitz exponent, the magnitude of which can be used as an indicator of the damage extent.Signal square exponent, a new index for damage identification is proposed based on discrete wavelet transform and Mallat algorithm. The proposed damage detection method is validated with a simulated simply supported concrete beam. Many aspects are analyzed in the simulated case, such as detecting the location of cracks, estimating the damage extent, anti-noise capability, and sensitivity to sensors' amount. The results show that the new candidate index achieved a good performance in the analysis and it is a validated and reliable index for damage identification.Multi-scale CWT based multi- resolution analysis possesses a good performance in the online structural health monitoring. A simulated simply supported concrete beam is employed for case analysis, and both cases with damping and without damping are taken into consideration. The results show that this method can detect the exact time when the damage occurs. And similar results are gained by dealing with signals from different sensors, which is of great use for the optimal disposal of the sensors and decreasing the amount of sensors.