Probing defects in friction stir welding process using temperature profile

Abstract

Detection of defects in friction stir welding process is a challenging task since most of the defects are internal or subsurface. An attempt has been made to explore the methodologies based on temperature signal for possible detection of defects in friction stir welding process using different tool profiles. The presence of defect is affected by temperature diffusion and is better reflected by the change of temperature over time. Temperature signals are acquired using thermocouples inserted in specific locations on advancing and retreating side of the welds. The rate of change of temperature and wavelet-analysis-based indicator computed from temperature signal against each experiment reveal appreciable difference for defective and defect-free welds. Threshold values are computed that clearly set a boundary for classifying the defective welds from defect-free welds. The proposed approaches can effectively reduce the post-processing steps essential for other non-destructive evaluation of the welds and can work as a first level of safeguard in the identification of defects during friction stir welding process. The methodologies can be extended towards monitoring of the process in industrial applications.