Skip to Main Content
The integrity diagnosis of tubular structures is studied in this research using guided acoustic waves. A smart sensor array is deployed to activate and collect Lamb wave signals that propagate along metallic tubes. Several advanced techniques have been explored to extract representative features from acoustic time series. Among them, the Hilbert-Huang Transform (HHT) is a recently developed technique for nonlinear nonstationary signal processing. A moving window is introduced to generate the local peak information from transient series, and a zooming window algorithm is developed to localize the structural flaws. The extracted features using moving windows are utilized for classifying structural defects in brass tubes. An experimental piezo-transducer system has been developed for testing several types of flaws in brass tubes in air and in water, and to demonstrate the implementation of HHT for flaw detection, localization, and classification. The techniques are also shown to be effective under background/process noise. This approach has the potential of on-line integrity monitoring of tubes in steam generators, boilers, and heat exchangers, used in process and power plants.