Application of Hilbert-Huang transform for ultrasonic nondestructive evaluation

In this investigation, the Hilbert-Huang transform (HHT) has been evaluated to characterize the ultrasonic backscattered echoes from materials with different grain sizes and defects. The HHT is a combination of empirical mode decomposition and Hilbert spectrum analysis. First, the ultrasonic signal is decomposed into a series of intrinsic mode functions. Then, based on the Hilbert spectrum of these intrinsic mode functions, a time-frequency representation of the ultrasonic signal is obtained. Furthermore, to demonstrate the application of HHT in ultrasonic signal processing, the performance of HHT has been compared with the results from other time-frequency techniques such as chirplet signal decomposition. Numerical and analytical results indicate that HHT is a unique and effective tool for ultrasonic signal analysis accounting for narrow-band, broad-band, and dispersive echoes. This algorithm can be utilized in the analysis of ultrasonic signals often encountered in flaw detection, signal classification, and pattern recognition.