Real-time defect detection in pulsed GTAW of Al alloys through on-line spectroscopy

A method for real-time monitoring of weld perturbations and defects in pulsed gas tungsten arc welding (GTAW) using the spectroscopy signal was developed. The statistic characteristic parameters of certain spectrum bands of interest (SOI), i.e., root mean square (R), variance (D) and kurtosis (K) are selected as the output monitoring parameters. 8 SOIs with different resolution are carefully selected, and the containing emission lines of chemical elements are identified and analyzed for seams with different disturbances. A correlation between the monitoring parameters and the seam defects is investigated, and the parameters D and K in SOI-7 (584.95–608.41 nm) have demonstrated the great sensitivity base on the SNR (signal to noise ratio) evaluation criterion. Wavelet packet transform (WPT) is carried out to remove the pulse interference in the monitoring curves. The proposed technique has been proved to be feasible in detecting the welding perturbation and seam defects.