Efficient feature extraction in ultrasonic spot weld inspection

Ultrasonic signals obtained in the process of nondestructive testing of resistance spot welds are often contaminated by various kinds of noise or artifacts. The data collection process in a robotic spot weld system occurs when the welding gun is fired and robot servo motors are running. This creates...

Full description

Saved in:

Bibliographic Details
Published in:	2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering (CCECE) pp. 1 - 4
Main Authors:	Baradarani, A., Khanli, L. M., Chertov, A. M., Regalado, W. P., Maev, R. Gr
Format:	Conference Proceeding
Language:	English
Published:	IEEE 01-04-2017
Subjects:	Feature extraction Inspection Noise reduction Robots Spot welding
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Ultrasonic signals obtained in the process of nondestructive testing of resistance spot welds are often contaminated by various kinds of noise or artifacts. The data collection process in a robotic spot weld system occurs when the welding gun is fired and robot servo motors are running. This creates very intense noise that can considerably diminish the ultrasonic inspection performance. Because of the nature and strength of noise in this case, key features such as the main pulse, first echo and backwall are substantially affected. In this paper, an efficient A-scan signal enhancement algorithm is proposed to extract desired features from highly contaminated signal mixtures. The algorithm is applied to the data collected using the industrial 6-axis ABB robot available at the Institute for Diagnostic Imaging Research and the data received from leading automakers in the North America and Europe.
DOI:	10.1109/CCECE.2017.7946670