Research On Properties Of Welding Joint And Weld Bead Dimension By Twin-wire Submerge Arc Welding Procedure

Owing to high deposition rate, excellent surface appearance, invisible arc, and lower welder skill requirement, the submerged-arc welding (SAW) process is used widely in the fabrication of pressure vessels, marine vessels, pipelines and offshore structures. Twin-wire submerged arc welding is one of advanced and efficient welding methods, the introduction of two-wire can reduce the welding pass and increase welding productivity obviously, and it can control welding heat input in a wide range to meet the performance requirements under different conditions by adjusting the pre-and post-wire welding current. In this paper, the effect of parameter on welded bead dimension was studied, in the hope that it can provide basis for actual production and reference for further research.Welding tests are carried out on the base metal 16Mn steel by single-wire direct current or alternating current or tandem sequence submerged arc welding technology using the MZS-1250 double-wire submerged arc welding machine. At the same welding rate and weld heat input, the deposition rate and the size of weld beads of the above-mentioned three welding technology are compared, the analyses on microstructure of heat-affected zone and weld metal are made, the mechanical property of welded joint are carried out. The results show that tandem sequence submerged arc welding technology can greatly increase the deposition rate and welding rate, the penetration and the width of weld obtained using the tandem sequence submerged arc welding technology are deeper and wider than that of single-wire direct or alternating current submerged arc welding technology, and the microstructure of heat-affected zone and weld metal is similar to that of single-wire direct current or alternating current submerged arc welding technology. Tensile strength and impact toughness of welded joints are slightly higher than that of the base metal.Weld geometric dimension and the structural performance determined by welding procedure parameters, so it is important to design welding procedure parameters for weld structure production. Based on numerical simulation and neural network, a BP neural network system of welding procedure parameters is built, in which welding procedure parameters are the input parameters, while weld quality information (weld geometric dimension) are the output parameters. Performance of this system is tested and validated. The simulation results indicated that the BP model has strong abilities in self-learning and predicting. The model possessed the ability to estimate the weld shape by the parameters.