Developing an adaptively-remeshed multiphysical resistance forge weld model, with extensive validation from residual stress measurement techniques

Understanding residual stress fields in manufactured components is crucial for predicting their performance under service loads; tensile residual stress can compromise strength, especially near stress concentrators or under cyclic loading conditions. This study introduces a multiphysical weld process model aimed at predicting the configuration and residual stress fields in a resistance-forge-welded port of a pressure vessel. Utilizing Lagrangian finite element analysis with adaptive remeshing, the model effectively addresses the significant deformations associated with welding. A comprehensive constitutive model is employed to capture material changes, including dynamic recrystallization during solid-state welding.

Model validation is achieved through displacement measurements and three residual stress estimation techniques: contour method, slitting method, and neutron diffraction, with a focus on hoop and axial stresses across three planes. The model’s configuration is further refined to align with destructive testing processes, enabling direct comparisons to measured residual stresses. Cross-validation of the contour method hoop-stress comparison was performed by simulating the contour cut and examining the relaxation displacement.

Results indicate a strong correlation between the model and measurements, particularly for axial stress from the slitting method, with a discussion on factors influencing discrepancies between model predictions and experimental data.