Numerical modelling is less frequently used in industry for simulation of welding than for simulation of other transformation procedures, like plastic deformation or melting. This is due to the multiphysical nature of the welding process, involving arc plasma, fluid flow in the melted area, strongly coupled mechanics, thermal effects and metallurgy. This complexity penalizes the setting up of innovative welding process, such as arc-laser hybrid welding. The industrial stake is big in sense of quality and productivity (no reliable predictive tool of the operative and metallurgical weldability exists). The aim of the SISHYFE project is to create this kind of tool, especially for the thick steel welding. Hence, four main aims are defined: 1. To develop methods of direct simulation of welding procedure, by modelling, in particular, laser-plasma interaction in case of hybrid welding, as well as the strong convective fluid flow of the liquid metal in the melted area. This should increase the predictability of these models. 2. In addition to direct simulation, to develop and to customize for hybrid welding (two power sources - arc and laser beam) a methodology consistent in identifying the thermal power sources using an automatic inversed finite elements method. 3. To evaluate the performance of these two methods, and particularly the contribution of the direct simulation, in sense of mechanical and metallurgical predictions (shape of the seam, properties of the melted area and the heat affected area, structure distortions and residual stresses). 4. To develop predictive simulations of the laser-arc hybrid welding procedure, that can also be applied to other procedures, such as arc, laser or electron beam welding. The methodology suggested is: A. Instrumented welding experiments performed for three configurations typical for hybrid welding will serve as reference throughout the project. Emphasis will be put on instrumentation with the aim to obtain a precise and consistent experimental database: thermocouples, high speed video camera, IR camera, distortion and deformation measurements using an image stereo-correlation device, ... B. For a development of numerical models three softwares will be used: COMSOL, SYSWELD and TRANSWELD. A verification of the numerical methods, on one hand, and the performances of each software, on the other hand, will be possible by comparing the obtained numerical results with the experimental results from corresponding reference configurations. C. These new softwares will be used for welding-tests by three industrial users. They would enable determination of experimental windows for operative and metallurgical weldability. The consortium is compact, reliable and well-balanced. It comprises all adequate expertises: 3 industrials, 1 technical center, 2 software companies and 2 research laboratories. All of them are major contributors to each of industrial, technical or scientific problematic. The project can contribute significant innovations in several fields of theme 4 of the request for proposals: - To develop innovative numerical tools for the comprehension of multiphysical phenomena typical for hybrid welding: plasma-laser interaction, fluid-structure coupling (hydrodynamics of melted area/solid part) - To verify the theoretical predictions using a serious experimental approach - To optimize the industrial development of the hybrid welding process