Objective In view of the welding deformation problems of medium- and thick-walled aluminum alloy profiles and plates in high-speed train carbody underframe, it is necessary to conduct numerical simulation analysis to comparatively study the heat input and deformation differences under different welding process parameters. This provides a theoretical basis and practical guidance for optimizing welding processes and controlling welding deformation, thereby improving the carbody manufacturing accuracy.

Method A three-dimensional simulation model is established using ABAQUS finite element analysis software. Based on the 2-layer 3-pass and 3-layer 6-pass welding processes used in actual production, a double-ellipsoidal heat source model is adopted to simulate the MIG (metal inert gas) welding process, and the pass-by-pass weld activation is realized through the ELM (element life and death method). Hexahedral meshing is applied to the model, with a mesh size of 1 mm in the weld zone and a maximum mesh size of 5 mm in the base material. A sequentially coupled method is employed in the numerical simulation: the nonlinear transient thermal analysis is first performed, and the resulting temperature field is then used as the initial condition for thermo-elastoplastic stress field calculation.

Result & Conclusion  For the 2-layer 3-pass process, the peak molten pool temperature reached 1 557.0 °C, significantly higher than the 879.3 °C obtained with the 3-layer 6-pass process. The maximum welding deformation of the 2-layer 3-pass process was 2.518 mm, markedly greater than 1.146 mm of the 3-layer 6-pass process, with a difference of 54.48% between them. The deformation difference is mainly attributed to the coupled effects of heat input, welding process parameters, and structural restraint. The 3-layer 6-pass process achieves a self-restraint effect by dispersing heat input and increasing inter-layer and inter-pass restraint, thereby effectively suppressing welding deformation. In practical engineering applications, when strict dimensional accuracy is required, the 3-layer 6-pass process should be preferentially adopted.