Monitoring of Weld Bead Profile in Friction Stir Lap Welding for Polycarbonate Sheets Using Wavelet Packets of Power Signal

The solid-state friction stir welding is found to be popular for the joining of thermoplastic materials in different configurations. However, the process variables have to be precisely adjusted with proper selection of tool pin profile to achieve uniform longitudinal weld along the tool traversing direction. The axial thrust and tool stirring torque signals are often used for the process monitoring and associated weld quality prediction. The present work addresses the effect of swept ratio of tool pin contour on the degree of non-uniformity in lap weld profile and index of symmetries in weld bead shape with associated mechanical behavior in tensile–shear loading for polycarbonate sheets. The hexagonal tool pin was found to be the best in terms of stability in weld contour, flawless symmetric weld bead shape with the highest strength efficiency (75.4%) and improved joint ductility particularly at high tool revolving speed (2400 rpm). The input power signal along with thrust–torque signals have also been acquired during welding procedure that was found to be chief indicator of weld quality. The electrical power loss was significantly less at higher tool rotational speed by using this hexagonal tool pin. Thus, this time field power signal has been further analyzed in time–frequency wavelet domain to completely characterize the process variations with corresponding deviations in weld profile. The high-frequency power wavelets were found to be highly sensitive on the deviations in weld quality features though low-frequency power information mainly dictated the variations in weld bead profile.