Baseline-Free Hybrid Diagnostic Technique for Detection of Minor Incipient Damage in the Structure

Abstract

One of the practical difficulties in implementing the structural health monitoring schemes in the field is the challenge involved in detecting the incipient minor damage (subtle cracks) at the earliest possible stage. The minor cracks alter only a few higher modes of vibration in a feeble intensity, which is more difficult to identify and isolate the damage information under the effects of environmental variability and noise in the overall signal. To handle this issue, a novel hybrid baseline-free damage diagnostic algorithm is proposed in this paper, combining an improved version of second-order blind identification (SOBI) with wavelets and time series models. An improved version of SOBI is used to identify and isolate the critical modal responses with damage features. Wavelet analysis on the critical modes identifies the time instant of damage. In a baseline-free framework, the current data is split into reference and damage data segments based on the time instant of damage, and so the environmental conditions are the same for both the segments. Thus, the difficulty of handling the environmental variability does not arise because the temperature variation between any two subsequent data segments is almost nil. The location of damage is identified using the time series analysis of the signal enhanced reconstructed data. Numerical studies and experimental studies are conducted to evaluate the capability of the proposed algorithm in identifying minor/incipient damage like small cracks with noisy measurements.