Key Engineering Materials Vols. 622-623

Abstract: A study was performed to evaluate how the Friction Stir Spot Welding process parameters affect both the thermal distribution in the welding region and the welding forces. An experimental campaign was performed by means of a CNC machine tool and FSSW lap joints on AA6060-T6 aluminum alloy plates having a thickness of 2+2 mm were executed. Five thermocouples were inserted into the samples at a specific distance from the specimen center. A set of tests was carried out by varying the process parameters, namely rotational speed, axial feed rate, plunging depth and dwell time. Axial welding forces were also measured during the execution of the experiments by means of a piezoelectric load cell. The experimental data collected were used to set up and to validate a simulative model of the process. In particular, a 2D FEM model was set up using the commercial code Deform 2D. A 2-dimensional FEM code was preferred in order to guarantee a very simple and practical model able to achieve results in a very short time. Since it is not possible to simulate the rotation of the tool in a 2D configuration, a specific external routine for the calculation of the developed thermal energy due to the friction between tool and workpiece was set up and implemented into the code starting from the local pressure distribution along the contact area.

Abstract: The recrystallization behaviour of a Twin Roll Cast AZ31 alloy, under deformation conditions corresponding to the initial stands of a magnesium hot rolling process, has been investigated. A set of strain levels, beginning at the critical strain for the onset of dynamic recrystallization has been studied. The dynamic recrystallization, taking place during hot deformation of magnesium alloy AZ31 is analyzed with help of plane strain compression tests conducted together with microscopic examination. Dynamically recrystallized volume fraction and size of dynamically recrystallized grains increase with increasing strain. When the strain is constant, the dynamically recrystallized volume fraction is higher at higher deformation temperatures or lower strain rates. Furthermore, it has been found out that deformation temperature plays a major role, rather than strain rates, in activating the dynamic recrystallization process during deformation. The results were used to identify the coefficients for the JMAK-model approach on the dynamic recrystallization of an AZ31 TRC-sheet.

Abstract: The effect of SiC addition on grain refinement of an AZ31 alloy was investigated. First preliminary casting tests were carried out. Different amounts between 0.25 wt% and 1,0 wt% of SiC-particles with a diameter of about 2 μm were added to an AZ31 melt and poured into a plate mould. The resulting microstructure was observed by optical microscopy. Best results were reached with the addition of about 0.25 wt% in terms of a homogeneous microstructure and a sufficient grain refinement. Higher amounts lead to coarsening and agglomeration of precipitations. Based on these results Twin-Roll-Casting (TRC) of an AZ31 alloy with the addition of SiC-particles was conducted. For the investigations a TRC-system in laboratory scale was used. This system is part of the Twin-Roll-Casting pilot plant at the Institute of Metal Forming, which is used in collaboration with the MgF Magnesium Flachprodukte GmbH. The produced TRC-strip was cut into plates, and samples for microstructure characterization and the determination of the mechanical properties were retrieved. Compared to TRC-strip without the addition of SiC-particles a microstructure with finer precipitations and a more homogeneous distribution can be observed. Furthermore, sheets with an addition of SiC offer improved mechanical properties in TRC-condition.

Abstract: In order to prevent the surface defect in the magnesium alloy extrusion process, it is important to set an appropriate process condition. The extrusion limit diagram is very useful to achieve the maximum extrusion speed without surface defect. In this study, the extrusion limit diagram for the magnesium alloy extrusion is constructed by using extrusion experiment and finite element analysis. For finite element analysis hot compression test is carried out to obtain the effective stress and stain curves according to the various strain rates and temperatures. The effectiveness of the constructed extrusion limit diagram is verified through the porthole extrusion experiment for producing the magnesium alloy bumper beam.

Abstract: The present investigation deals with the development of a methodology to predict the flow behaviour of the ZM21 magnesium alloy in given intervals of temperature and strain rate by FEM simulation of torsion testing. Equations based on the hyperbolic sine of flow stress and on the multiple linear regression were proposed and implemented into the finite element code. The flow curve shapes obtained by simulation were compared with experimental ones that were not used in the building phase of the equations. It was found that the simulation of torsion tests allows, under given conditions of temperature, strain rate and deformation levels, to obtain flow curve shapes very similar to those obtained by experiments under conditions not included in the building of the models.

Abstract: Interest in lightweight materials, particularly magnesium alloys, has increased significantly with rising efficiency requirements in the automotive sector. Magnesium is the lightest available structural metal, with a density approximately 35% lower than that of aluminium. The potential is great for magnesium to become a primary material used in future low carbon vehicle structures; however, there are significant obstacles, namely low ductility and formability, particularly at room temperature. The aim of this work is to present the feasibility of using the solution Heat treatment, Forming, and in-die Quenching (HFQ) process to produce complex shapes from a sheet magnesium alloy, and to use the results to verify a simulation of the process developed using commercial FE software. Uniaxial tensile tests were initially conducted to establish the optimum parameters for forming the part. Stamping trials were then carried out using these parameters, and a simulation set up modelling the forming operation. It was shown that the HFQ process could be used to form a successful component from this alloy, and that a good match was achieved between the results of the forming experiments and the simulation.

Abstract: The objective of this study was to understand the work-hardening behavior of a rolled AZ31 magnesium alloy sheet upon reverse loading. Firstly we carried out an in-plane compression-tension test of a rolled AZ31 magnesium alloy sheet with various compressive strains. The sigmoidal curve was exhibited during tension regardless of the amount of compressive strain, but a shape of the curve was clearly different depending on the compressive strain. To understand the mechanism of this difference, a crystal plasticity finite-element simulation was carried out. The simulation result showed that the above difference in the shape was owing to the difference in the activities of slip and twinning systems during tension depending on the compressive strain.

Abstract: In this paper, finite element predictions of a tensile test of cylindrical specimens obtained by rigid-plastic and elastoplastic finite element methods are compared in terms of tensile load-elongation curve and deformed shape. The flow stress curve used for this study is obtained by a scheme of obtaining flow stress at large strain from tensile test of cylindrical specimen using rigid-plastic finite element method. The two predictions are compared in a quantitative manner and discussed not only to find some similarity but also to distinguish the elastoplastic finite element method from the rigid-plastic finite element method.

Abstract: Development of a numerical model for the dynamic recrystallization (DRX) on the basis of the Cellular Automata method and the Digital Material Representation (DMR) idea is the main goal of the present work. Basic assumptions (space definition, neighborhood type, state and internal variables) of the proposed model are presented and discussed. Particular attention is put on description of the developed transition rules used to replicate mechanisms leading to dynamic recrystallization. Finally, examples of obtained results of DRX morphology and kinetics are also presented within the paper.

Abstract: The 3D finite element simulations of processes like multi-pass rolling or drawing often result into exorbitant computational times that make the numerical approach almost infeasible while only the “stationary” step of the process is actually of interest for the industry. Therefore, it can be advantageously simulated by resorting to steady-state formulations which allows reducing the calculation time by, at least, an order of magnitude with respect to more conventional methods where the steady regime is incrementally calculated. A general and robust formulation is developed; it is suitable for parallel computing, compatible with unstructured meshes and general enough to apply to a wide range of forming processes. It consists in alternatively resolving the “simple” steady-state material forming problem for a given domain geometry and then computing the domain corrections that allow satisfying the free surface condition. Within the “simple forming problem”, a Streamline Upwind Petrov Galerkin (SUPG) method is used to integrate the state variables along the streamlines. For the domain geometry correction, a Least Squares formulation with an Upwind shift is introduced. The two resolutions are coupled by the contact equations. This method is applied to several metal forming problems such as 3D rolling and drawing. Results show the high efficiency of the method with respect to an incremental resolution. Computational time is reduced by a factor ranging between 20 and 30. Results are as accurate as with an incremental method. Convergence is always reached whatever the initial geometry of the domain at the beginning of the iterative algorithm.