Ultrasonic Detection of the Metallurgical Defects in the Steel and its Evaluation by Neural Network Based on the Wavelet Transform Noise Suppression

Abstract

To detect the metallurgical flaws and to realize its automatic classification has long been expected by technicians in the steelmaking shop, among various methods have been tried, ultrasonic is the mostly preferred one. In this work, a method used to realize the steel flaws detection and auto-classification has been tried, which involving the use of the high center frequency ultrasonic detection, transient nonstationary signal processing tools, wavelet transformation, and neural networks. On the base of ultrasonic test results, the different flaw echoes were analyzed by multiresolution signal decomposition using wavelet transform, and a chain of 120 data samples were extracted from the reconstructed wavelet coefficient, which can represents the character of the corresponding flaw. Totally there were 42 chains of data were collected from specimens with different flaws contained, they were divided into a training set and a testing set for the training of the neural networks, which are used for automatic classification of the flaw signals. The results produced from the network architectures investigated shown classification rates in the range from 70 to 100% based on real and artificial defects and with comparatively simple feature extraction method, a performance can be accepted in the light of rough evaluation

Crack identification in plates using 1-D discrete wavelet transform

“…acoustic emission (Rogers, 2005), X-rays (Shinoba et al, 2004), eddy current (Gros, 1995), ultrasonography (Zhang et al, 2004), magnetic field (Lee et al, 2004) or soft computing methods such as artificial neural networks (Waszczyszyn and Ziemiań-ski, 2001) and evolutionary algorithms (Burczyński et al, 2004). The traditional approach, based on the analysis of natural frequencies (Dems and Mróz, 2001) and modal shapes of structure vibrations (Ostachowicz and Kaczmarczyk, 2001), is still used and further developed.…”