Numerical Modelling of Laser Beam welding of Aluminium Alloy-Steel Joints

A significant trend is noticed in automotive industry towards a wider use of aluminium and aluminium alloys (AA). They have great specific strength and a high corrosion resistance allowing the design of lightweight constructions with excellent mechanical properties. In this context, the demand for laser welding of the material combination of aluminium alloys with steel (AA/St) has increased. However, laser beam welding (LBW) desired for longer AA/St joints or more complex joints (e.g. car door/hinge or multiple sheet structures in car bodies) is still difficult on account of the formation of extreme brittle compounds at the intermetallic interface. Controlling and predicting this formation is at present very difficult due to the current lack of understanding of the involved solidification and transport phenomena in the weld pool.The aim of the project is the development of a numerical model of deep penetration laser beam welding of the material combination steel/aluminium alloys. It will allow the prediction of:· thermal cycle in fusion zone and intermetallic interface· formation of intermetallic phase (brittle compounds AlxFey)· shape of the seamA completely new approach will be applied to evaluate the solidification phenomena at the interface phase between the aluminium and steel part considering the complex multi-phase heat transfer problem and the fluid flow of a binary mixture. The basis of the algorithm will be a finite volume discretisation of the conservation equations of energy, mass and momentum, of the minimisation of surface energy and of the thermo-metallurgical phenomena. The extreme fine spatial discretisation necessary to resolve the intermetallic layer will be realised employing a composite grid technique.Although the project has a focus on the industrial aspect, the main content of the project is fundamental research at high scientific level. The development of the model describing the physics of the complex non-linear problem and of an efficient numerical algorithm for the underlying large-scale discretisation is the main task of the project. Considering the complexity of the problem, it cannot be expected to reach the development of a complete industrial simulation tool within the two years frame of the project. But it is expected that the improvement of the understanding of the phenomena will make a sound foundation for such a development.