Baseline-free damage identification method based on guided wave for tube structures

Abstract The tube structure is a crucial component in petrochemical industry or other chemical plants, any damage could lead to severe consequences. With the widespread adoption of guided wave technology in structural health monitoring, this study presents a novel guided wave-based, baseline-free approach for localizing damage in tube structures. The core concept of this methodology involves constructing a comprehensive guided wave pattern encompassing all potential guided waves within the tube structure using the Gabor pulse model, which adeptly captures both the frequency and time attributes of guided waves. Subsequently, the composition of wave packets within each receiving sensor is determined utilizing the block sparse Bayesian learning algorithm. By optimizing the sensor arrangement strategy, a damage wave packet identification algorithm is proposed to discern the damage wave packets effectively. This approach enables damage recognition without relying on traditional baseline data by utilizing parameters from the Gabor pulse model within the damage wave packet combined with probability density. Through rigorous evaluation utilizing simulation and experimental data, the efficacy of the baseline-free damage identification method is demonstrated, with the maximum error in damage prediction confined within 8 mm.