Numerical welding simulation of a vessel head adapter - Influence of modelling assumptions

In nuclear industry, most of heavy components and reactor coolant lines have a large thickness and their manufacturing processes require multi-pass welding. When low-alloy steel components are concerned, the assembly process is often performed in several stages, such as a cladding, a buttering and a Post Weld Heat Treatment (PWHT) before joining two materials without phase transformations. The distortions induced by the welding operation might be an issue and residual stresses could be significant and play a role on the weldability. For these reasons, AREVA has placed a lot of effort to improve the reliability of numerical simulation of its welding processes, in order to have a better understanding of the involved phenomena and also to predict the residual stress state through the structure [1], [2] and [3] because this numerical simulation can be used to select the manufacturing process in the early phase of welded component design. The aim of the simulations presented in this paper is the investigation of the final residual state of a nozzle placed at the central position of the vessel head. The computations are performed according a robust methodology packaged in AREVA OSS tool [4] which is based on SYSWELD™ Finite Element solver [5]. Two welding configurations are investigated. The first one is a mock-up with an "open" narrow gap, the groove filling being performed by a manual welding process using electrode coatings. The second one is a mock-up with a "closed" narrow gap, the groove filling being performed by an automatic TIG process. After the comparison of these two configurations, a special investigation is performed on the "open" narrow gap mock-up. The influence of the vessel head buttering before the groove filling is investigated, as well as the efficiency of the PWHT performed after the buttering operation.