Two-dimensional (2D) and three-dimensional (3D) weld-induced residual stress finite element analyses have been performed for 2-inch Schedule 80 Type-304 stainless steel pipe sections joined by a multi-layer segmented-bead pipe weld. The analyses investigate the similarities and differences between the two modeling approaches in terms of residual stresses and axial shrinkage induced by the pipe weld. The 2D analyses are of axisymmetric behavior and evaluate two different pipe end constraints, namely fixed-fixed and fixed-free, while the 3D analysis approximates the non-axisymmetric segmented welding expected in production, with fixed-free pipe end constraints.
Based on the results presented, the following conclusions can be drawn. The welding temperature contour results between the 2D and 3D analyses are very similar. Only the 3D analysis is capable of simulating the non-axisymmetric behavior of the segmented welding technique. The 2D analyses yield similar hoop residual stresses to the 3D analysis, and closely capture the maximum and minimum ID surface hoop residual stresses from the 3D analysis. The primary difference in ID surface residual stresses between the 2D fixed-fixed and 2D fixed-free constraints cases is the higher tensile axial stresses in the pipe outside of the weld region. The 2D analyses under-predict the maximum axial residual stress compared to the 3D analysis. The 2D ID surface residual stress results tend to bound the averaged 3D results. 2D axisymmetric modeling tends to significantly under-predict weld shrinkage. Axial weld shrinkage from 3D modeling is of the same magnitude as values measured in the laboratory on a prototypic mockup.