Ultrasonic testing of internal defects in welding seam of steel pipe pile in port terminal bent structure

In order to overcome the problems of low defect location detection rate, low defect type detection accuracy, and long detection time in traditional defect detection methods, an ultrasonic testing method of internal defects in welding seam of steel pipe pile in port terminal bent structure is proposed. Using an ultrasonic flaw detector to collect ultrasonic signals from steel pipe pile welds, the CEEMD algorithm is used to denoise the ultrasonic signals and extract signal features. The extracted features are input into a support vector machine, which introduces Karush-Kuhn-Tucker (KKT) conditions in optimisation theory and Lagrange multipliers to find the hyperplane with the maximum spacing. This hyperplane is used to achieve ultrasonic detection of internal defects in the weld. Experimental results show that the maximum defect location detection rate of the proposed method is 98.76%, the maximum defect type detection accuracy is 98.69%, and the detection time varies between 0.21 s and 0.39 s.