Prediction of Weld Bead Geometry in Saw Using Regression Method

Submerged Arc Welding (SAW) is a metal joining technique widely used in heavy fabrication industries due to its inherent properties. The welding quality and productivity is controlled by the process parameters. The process planners use different techniques to estimate the influence of the welding parameters (welding current, arc voltage, welding speed and electrode stickout) on bead geometry. This paper discusses about the design of experiments and the development of Multiregression model to predict the weld bead geometry. The developed model determines those variables which will give the desired set of bead geometry (weld bead width, weld reinforcement, weld bead penetration, reinforcement area, and area of penetration and bead dilution).