Development of Weld Modelling Guidelines in the UK

This paper describes the development of finite element modelling guidelines for the calculation of welding residual stresses. These guidelines form a new section in the R6 procedure, used in the UK nuclear power industry for the assessment of integrity structures containing defects. The intention is to improve the consistency of weld modelling procedures, the accuracy of predicted residual stress profiles and confidence in their use for defect tolerance assessments. The first issue of these guidelines is applicable to austenitic stainless steel joints produced using arc welding processes. The components of interest are mainly thick section nuclear pressure vessels and pipe welds where distortion is not the key issue. Recommendations made in the guidelines are largely based on residual stress analysis methods, validated by measurements on a range of weld mock-ups, developed over several years in support of British Energy projects. Advice is included on the use of 2D and 3D models, welding heat sources, material properties requirements, cyclic hardening and annealing assumptions. The modelling and computational requirements depend on the level of accuracy and degree of validation required. This is likely to be a function of the defect tolerance in the structure. In future issues, the R6 modelling guidelines will be supported by weld validation benchmarks. This will provide a detailed manufacturing record and measurement data from controlled weld mock-ups (including specimen design, welding parameters, thermo-mechanical properties, thermocouple data and stress measurements). It is also planned to develop these guidelines to include ferritic steel and dissimilar metal welds. The metallurgical behaviour in ferritic steel welds is more complex, since micro-structural phase transformations occur. Guidance will be provided on modelling post-weld heat treatment (PWHT) applied to pressure vessel welds and stress relaxation by creep. In modelling dissimilar metal welds, it is necessary to provide advice on dealing with the structural discontinuity at material interfaces and overcome FE solution convergence problems.