Defect classification of electrofusion joints in polyethylene pipes

 

FU Xibin1, HUANG Yuexin2, HUANG Xuebin1, ZHONG Shuncong2,3

 

(1. Xiamen Special Equipment Inspection Institute, Xiamen 361004, China;

 

2. School of Mechanical Engineering and Automation, Fuzhou University, Fuzhou 350108, China;

 

3. Fujian Key Laboratory of Medical Instrument and Pharmaceutical Technology, Fuzhou 350108, China)

 

 

Abstract: In order to classify void defects and heating wires in electrofusion joints of polyethylene pipes, defects′ location and barycenter were obtained from ultrasonic phased array images using image processing methods. Onedimensional signals, the gray scale values of the column signal at the locations with different defect barycenter in ultrasonic phased array images, were calculated accordingly. Highfrequency components of these onedimensional signals were achieved by using wavelet decomposition algorithm. Experimental results demonstrate that statistical behavior of the highfrequency components can be effectively used in defect classification, especially for the void defects and heating wires. Therefore, this defect classification strategy can be recommended in real engineering applications.

 

Key words: defect classification; defect barycenter; wavelet decomposition; statistical behavior