Research on Multi-layer & Multi-pass Welding Processes Domestically and Internationally

Research on multi-layer and multi-pass welding processes, both domestically and internationally, primarily focuses on welding residual stress, with an emphasis on predicting and eliminating residual stress distribution post-welding. Research methods can be categorized into two approaches: experimental research and numerical simulation. Heat treatment is commonly used to eliminate residual stress after welding. However, due to the inherent structure of welded components, not all parts can undergo heat treatment to eliminate residual stress.

To enhance the overall performance of the forged valve, the mechanical properties and metallographic structure of vibration multi-layer and multi-pass submerged arc welds are compared with those of ordinary multi-layer and multi-pass submerged arc welds. The analysis shows that the grain size in the metallographic structure of vibration multi-layer and multi-pass submerged arc welds is smaller than in ordinary welds. In terms of mechanical properties, the bending and impact performance of vibration welds are superior.

To further eliminate post-weld stress in valve body welding, Jijin Xu  proposed a mechanical stress elimination method. This method applies external pressure to the welded structure. The residual welding stress interacts with the external pressure, inducing plastic deformation in the welded parts and thus reducing the stress. A loading pressure and time relationship curve was developed for this method, offering a new approach to stress elimination in welded parts without heat treatment.
 
In experimental research, Qiang Bin measured the residual stress distribution of Q345 steel welded joints, showing that the residual stress distribution across different thicknesses is consistent. Li and others measured the surface stress of Q690 and Q460 steels using the strip cutting method, showing that the surface stress of Q690 steel is relatively low. Peddea used the neutron method to measure the residual stress after steel pipe welding and found tensile stress concentration on the inner surface and compressive stress on the outer surface. Other researchers have employed various non-destructive testing methods to measure welding residual stress.

In numerical simulation research, Ueda and Hibbitt developed a thermo-elastic-plastic finite element method to calculate welding stress. Goldak's double ellipsoid heat source model improves the accuracy of welding temperature field calculations. Bhatti and others studied the influence of material parameters on welding residual stress, noting that specific heat and thermal conductivity significantly impact the calculation results, while density has a lesser effect. Advancements in computer technology have significantly improved welding finite element simulation.

Dae-Won Chol studied the effects of welding gun angle and current polarity on heat exchange, developed an arc heat flux model, and demonstrated the influence of arc force, droplet polarity, and other factors on welding seam formation. Studies show that phase changes during welding significantly impact residual stress, and simple thermal-mechanical coupling analysis no longer fully reflects welding stress changes. Leblond studied the effects of diffusive and non-diffusive phase changes on welding residual stress. Deng and others considered phase changes during welding when predicting the residual stress of 9Cr-1Mo steel, which significantly improved calculation accuracy. Other researchers have explored the effects of solid-state phase changes in welding on residual stress in different materials.

In summary, the study of welding residual stress relies not only on experiments but also on finite element numerical simulation. Currently, accounting for metallurgical effects in welding is the focus of residual stress research.

The optimization of the welding process involves optimizing both the welding method and the welding parameters. When the welding method is fixed, optimizing the welding parameters is crucial for improving welding quality. Scholars, both domestically and internationally, have conducted extensive research to clarify the influence of welding parameters on welding quality. Dhas summarized the steps for optimizing welding parameters, Holub investigated the effect of different welding wire metals on the strength of welded joints, and Ghosh examined the influence of welding parameters on weld quality.

The orthogonal test method and the mathematical statistics method are widely used in welding parameter optimization. For example, Yang used the orthogonal test method and fuzzy logic technology to optimize welding parameters, while Muhammad applied the orthogonal test method to optimize submerged arc welding parameters, combining it with the signal-to-noise ratio analysis to determine key parameters. Guodong Zhang and others analyzed the influence of welding parameters on welding stress through orthogonal tests and finite element calculations. The use of intelligent algorithms in welding parameter optimization is also increasing. D.S. Nagesh used a neural network model to predict weld shape and combined it with a genetic algorithm to optimize welding parameters. Guo Yu and others embedded the orthogonal test results into a BP neural network, established a prediction model, verified its accuracy, and provided new ideas for welding process optimization.

In summary, whether through mathematical statistics, finite element calculations, neural networks, or genetic algorithms, the optimization of welding parameters relies on experimental data. By combining orthogonal experiments with mathematical statistics, welding parameters can be effectively optimized, improving the quality of the welding process.