In the car manufacturing industry the actual main goal to have a better fuel economy with a minimal CO2 environmental impact requires to use lighter weight materials having better mechanical properties and possible recycling processes. It is the principal reason than the most used metallic materials in the automotive industry are the aluminium alloys and the dual phase (DP) steels. Concerning the assembly process of car body parts the welding ability together with a high fatigue life strength properties are critical to their successful exploitation. Laser welding assembly technology is today used widely in many industrial fields such as spaceflight, aeronautics, naval and automotive production. It is generally characterized by low heat input, high welding speed, high penetration, easy automation, high accuracy, fast and robotic production lines.

Laser-welded structures are often subjected to dynamic service loadings ranging from cyclic fluctuations to completely random ones. Unfortunately the laser-welded lap joints suffer from defects resulting in the notch effect and surface cracks, residual strains and residual stress, etc. The fatigue strength of the laser-welded lap joints is reduced significantly because the presence of these defects. Therefore, the mechanical strength of laser welded structures must be defined in terms of the fatigue strength and residual stress of the obtained joints or assemblies. Furthermore fatigue failures occur when metal is subjected to a repetitive or fluctuating stress and will fail at a stress much lower than its tensile strength or stress-strain behaviour.

In order to analyse the above discussed effects this scientific paper proposes an experimental and numerical analysis of elasto-plastic fatigue strength using uni-axial cyclic tensile tests (with N = 10000 cycles) of assemblies obtained from overlapped thin DP 600 steel sheets. The presented results in terms of fatigue or strain-life curves, strains and residual stresses spatial distributions obtained at the end of a low fatigue cycle were taken from an Abaqus Finite Element Modelling using a four-node linear tetrahedron C3D4 mesh and simulations with a kinematic hardening model and cyclic amplitude displacement boundary condition of max 16 mm. In order to compare robustness of a DP600 steel's Nd:YAG laser-welding, three different low cyclic fatigue tensile configurations are analysed: a simple thin sheet with a thickness of 1.25 mm, a perfect or ideal welding assembly of two identical sheets without any gap and an laser welding sheet's assembly with a gap of 0.1 mm. The obtained results show that after the cycling tensile loadings, the estimation of residual stresses on the central welded specimen area has for the laser-welding assembly lower values as compared to the simple specimen and relatively close to the named perfect or ideal assembly case. It is then possible to conclude on the good reliability of the of the used laser beam welding process.