A time-frequency method for ultrasonic flaw inspection based on HHT

A time-frequency analysis method based on Hilbert-Huang transform (HHT) was employed to character ultrasonic flaw echo signals. In this method, time and frequency characteristics of ultrasonic transducers used in inspection were considered to eliminate noises and identify the flaws. First, the flaw echo signals were decomposed into a series of mono-components called intrinsic mode functions (IMF) using the theories of EMD. Secondly, the best IMF was identified and represented in time-frequency domain using Hilbert transform, time and frequency characteristics were used to determine the flaw. Finally, the addition of the best IMF and the hypo-best IMF comes close to approximating the original flaw signal. The results of experiments with three transducers validate the effectiveness of the method in ultrasonic echo signal analysis.