Fuzzy Logic Approach to Modeling Weld Bead Geometry in ARC Welding

Weld joint quality depends on welding process settings. High-quality welds need controlled input settings. This study looks at four input settings: arc speed, wire feed to travel speed ratio, wire feed rate, and eccentricity. Tests used a full factorial design on a mild steel joint. Bead penetration, height, and width were measured. Fuzzy logic created models for these output measures. This fuzzy model predicts the output settings. Also, weld shape accuracy was checked. The inaccuracy is below 20% for bead penetration. Bead height and width inaccuracy is usually less than 10%.