Heat source modeling, penetration analysis and parametric optimization of super spray MAG welding

Abstract

Main drives, cutterheads and other critical components of tunnel shield machines require welding with thick plates that leave roots over 5 mm. Full penetration welds cannot be achieved by conventional Pulsed MAG welding methods. This article introduces Super Spray MAG Welding technology and investigates its penetrating regularities and mechanisms through high-speed camera images, finite element simulation, and microstructural analysis. An optimal welding procedure was generated using a combination of Genetic Algorithm and Back Propagation Neural Network. The data show that Super Spray MAG arc exhibits greater concentration and stability than traditional MAG arc, marking its strong qualities in emitting high-energy beams. The morphological solidification pattern of the molten pool closely matches the FEM simulation results of the composite Gaussian surface heat source model and peak linear attenuation Gaussian cylinder heat source. The welding current mainly affects the penetration of the weld, followed by the extension of the wire, and lastly the welding speed. Increasing the welding current can transition droplet transfer from globular to spray, as well as alter microstructure development and mechanical characteristics. Suggested parameters for penetrating the 5 mm root were put forward. The BPNN-GA model established can effectively predict weld formation, and points out the optimal welding parameters.