Tool Quality Monitoring In Friction Stir Welding Process

The present research aims at monitoring the quality of the tool in an advanced joining technique named friction stir welding by analyzing the force signals. Five tools with different qualities, namely a cracked tool, a tool with a half-pin, material sticking to the pin, tool without a pin, and a healthy tool, have been selected. Discrete wavelet transform has been applied on the acquired signal, and statistical feature of the wavelet coefficients has been extracted which aid in differentiating the selected tools' condition. The depicted idea will improve the weld quality, reduce rejection of product, and subsequently will help avoid machine downtime which may occur because of a faulty tool.