Prediction of Weld Bead Geometry for Robotic GMAW Rapid Prototyping & Manufacturing
This paper concentrates on direct rapid prototyping and manufacturing (RP&M) of functional metallic parts. Robotic gas metal arc welding (GMAW) is employed in this “Slicing & Stack” principle RP&M system. It is indicated that surface smoothness is a critical factor to affects the performance of RP & M products. In order to improve surface smoothness of product, the RP & M system must decrease stack error during stacking in each layer. This investigation establishes relationships between welding parameters and weld bead geometry. First, a rational welding parameters range is determined according to preliminary experiments. Then, quadric orthogonal regression rotational combination experiments scheme is proposed to predict width and height of weld bead. The width and height in regression results are expressed in the form of quadratic equations by welding parameters. Significance test results show that the two quadratic equations are both significant. According to the established relationships, users can easily predict width and height of weld bead when welding parameters are given. Whereas when the given condition is weld bead geometry, optimum welding parameters can also be determined by importing boundary condition according to users’ requirement or the service environment of parts. Experiment results indicate that prediction errors of width and height are both less than 3%.