Modeling of laser beam welding of stainless steel sheet butt joint using neural networks

Abstract

Laser beam welding (LBW) of austenitic stainless steel (AISI 304) of thickness 1.6 mm is carried out using 2 kW CW Nd:YAG laser. The effect of laser power, welding speed and beam incident angle on the butt weld bead geometry, i.e. depth of penetration (DOP) and bead width (BW) is investigated. The experiment is designed based on three level Box-Behnken design with replication. Modeling of laser welding of stainless sheet butt joint is performed using backpropagation trained neural network. The predicted results of the research showed a comprehensive and usable prediction of the laser welding parameters for butt joints using neural network.