Defect detection for automated and robotic gas metal arc welding systems using wavelet analysis

Gas Metal Arc Welding (GMAW) is a joining technique which is widely used by industry. The advantages of this process when compared to other joining techniques are its reliability, strength and the low cost of consumables. The necessity for automated or robotic welding systems arieses from the need to improve productivity and quality as well as avoiding the exposure of human operators to hazardous environments. These welding systems have a large number of applications in fields such as the construction of large structures, automotive and aeronautic, industries, and even microelectronics [1]. Nowadays, the industrial market is mainly driven by the need to increase manufacturing productivity, flexibility, performance and the prospect of decreasing purchase price and operating cost per unit throughput [2]. At the present time, detection of defects in GMAW usually involves either destructive or non-destructive methods, both of which are post-operative and costly. Through the arc sensing techniques might constitute an effective alternative means of detecting defects on-line [3].