LAUSR

In this study, a simulation-based examination on the deformation mechanism in the friction stir welding (FSW) process is conducted, which may not be efficiently feasible by experiment due to severe deformation and rotation of material flow near a tool pin. To overcome the severity of distortion of plastically deforming finite element meshes in the Lagrange formulation, and an over-simplified elastic-plasticity constitutive law and contact assumption in the Eulerian formulation, the arbitrary Lagrangian–Eulerian (ALE) formulation is employed for the finite element simulations. Superior accuracy in predicting the temperature profiles and distributions of the friction stir welded aluminum alloy workpiece could be obtained compared to the results of Eulerian based simulations. In particular, the ALE based simulations could predict the sharper gradient of temperature decrease as the distance from the welding zone increases, while the Eulerian based model gives more uniform profiles. The second objective of the study is to investigate the coupling of simulation-based temperature histories into the strength prediction model, which is formulated on the basis of precipitation kinetics and precipitate-dislocation interaction. The calculated yield strength distribution is also in better agreement with experiment than that by the Eulerian based model. Finally, the mechanism of the FSW process is studied by thoroughly examining the frictional and material flow behavior of the aluminum alloy in the welded zone. It is suggested that the initially high rate of temperature increase is attributed to frictional heat due to slipping of material on the tool surface, and the subsequent saturated temperature is the result of sequential repetitive activations of the sticking and slipping modes of the softened material. The sticking mode is the main source of plastically dissipated heat by the large plastic deformation around the rotating tool pin. The present integrated finite element simulation and microstructure-based strength prediction model may provide an efficient tool for the design of the FSW process.