Development of a Residual Stress Simulation Benchmark for a Single Bead-on-Plate Weldment

Finite element weld residual stress modelling procedures involve complex non-linear analyses where many assumptions and approximations have to be made by the analyst. Weld modelling guidelines for inclusion in the R6 defect assessment procedure are in preparation and will be accompanied by a series of validation benchmarks that can be used to evaluate the accuracy of weld modelling procedures and assess their suitability for use in fracture assessments. It is intended to base one of the benchmarks on a stainless steel bead-on-plate weldment that has been extensively studied by members of Task Group 1 of the NeT European Network project. This paper uses round robin residual stress measurements from the NeT project to derive a statistically based ‘best estimate’ distribution of transverse stress passing through the wall-section at mid-length of the bead-on-plate weldment. The accuracy of a state-of-the-art residual stress prediction is benchmarked against the best estimate measurements using a root mean square error analysis and comparisons of decomposed components of stress. The appropriateness of using the predicted residual stresses in fracture assessments is assessed by comparing stress intensity factors based on the measured and predicted distributions of stress. The results from these studies will be used to help establish accuracy targets and acceptance criteria for the welding benchmark.