Excerpt from the January 2010 issue of Metallurgical Science & Technology by Giulio Timelli
The increased use of light alloys in the automotive industry is, above all, due to the need of decreasing vehicle weight. The same need has to be taken into account in order to meet both energy and environmental requirements. In terms of application rates, Al and its alloys have an advantage over other light materials, such as Mg and Ti alloys. The reduced prices, recyclability, development of new improved alloys, increased understanding of design criteria and life prediction for stressed components and an excellent compromise between mechanical performances and lightness are the key factors for the increasing demand of Al alloys.
A great contribution to the use of Al alloys comes from improvements in casting processes, which allow to increase the production, to reduce the cycle time, and to manufacture complex-shaped castings with thin wall thickness. Among the recent casting techniques, the high-pressure die casting (HPDC) is largely used by the automotive sector since it fulfills the above advantages.
A limit to large diffusion of HPDC remains the final quality of castings. While the combination of high speed casting and high cooling rate gives the possibility of thin walled castings, the associated turbulence remains the major source of inner and surface casting defects, which have a deleterious effects on mechanical properties. In HPDC if a number of parameters are not adequately determined and adjusted, the quality of the die cast part results are rather poor. Macro-segregation of eutectic, primary inter-metallic particles and α-Al crystals, porosity, oxide bi-films and confluence welds are addressed as typical HPDC defects.
By means of casting parameters’ adjustments, foundrymen try to restrict and isolate the major part of defects into regions of the casting which are not mechanically stressed during normal working. Further, thin-walled castings, like those produced by HPDC, are more affected by the presence of defects since a single macro-defect can cover a significant fraction of the cross-section area. Many researchers have investigated the influence of casting defects on the mechanical properties both of gravity-cast and high-pressure-die-cast aluminum alloys. However, the works of Gokhale et al.  and Timelli et al.  show how the mechanical properties decrease monotonically with increasing the area fraction of defects revealed on the fracture surfaces both of gravity and high pressure die cast aluminum specimens. The common conclusion was that even high integrity castings contain defects, and thus, it is important to predict their effect on final mechanical properties.
Constitutive Model of Tension Instability
When porosity or an equivalent defect is present in a tensile specimen, the load bearing area is reduced. Thus, the defective region will yield first, concentrating the strain. The rate of strain concentration can be calculated considering the strain hardening ability of the material. A geometric defect that locally reduces the load bearing area of a tensile specimen results in the formation of an incipient neck. The growth of this neck can be described using the Ghosh constitutive model for the development of plastic instabilities. If the neck is not sharp or the strains involved are not large, it may be considered that only one significant stress exists in either the uniform section or the local inhomogeneity. Under the assumption that the material containing internal discontinuities experiences a tensile load under axial local equilibrium and the effects of strain rate can be neglected, the conventional equation for stress distribution can be expressed in terms of load carrying area as σi (1 - f)A0e-εi= σhA0e-εh (1) where σi , εi and σh, εh are the true stresses and strains in and outside the defect region, respectively, A0 is the initial cross section of the specimen and f is the area fraction covered by defects.
The secondary AlSi9Cu3(Fe) cast alloy (EN AC-46000, equivalent to the US designation A380) was supplied as commercial ingots, which were melted in a 500 kg SiC crucible in an electric resistance furnace. Before pouring the melt is held in the furnace at 690 ± 5°C for 1 h to ensure homogeneity and dissolution of the present intermetallic. Periodically, the molten metal was manually skimmed and stirred with a coated paddle to avoid any type of sedimentation. The furnace temperature is the holding temperature commonly used for EN AC-46000 type alloys, which is enough to avoid sludge formation [34,35]. The chemical composition, measured on separately poured samples, is shown in Table 1. For R&D purposes, a die for U-shaped casting was made.
Table 1: Chemical composition of the alloy (wt.%)
The CAD model of Al casting with runners, gating and overflow system is displayed in Figure 1. The U-shaped casting with 2.5 mm thickness was coupled with ribs, with ~5 mm thickness, which are generally locations of high defect content. The castings were produced in a Müller-Weingarten cold-chamber die casting machine with a locking force of 7.4 MN. The weight of the Al alloy casting was 3.3 kg, including the runners, gating and overflow system. A detailed description of the HPDC machine, the casting procedure, and the process parameters is given elsewhere. Briefly, 10 to 15 castings were scrapped after the start up, to reach a quasi-steady-state temperature in the shot chamber and die. Oil circulation channels in the die served to stabilize the temperature (at ~230°C). The fill fraction of the shot chamber, with a 70 mm inner diameter, was kept at 0.56. A combination of injection parameters and pouring temperatures were chosen in order to generate different types and amount of casting defects.
Result and Discussion
Analysis of Fracture Surfaces
It was found that the U-shaped castings contain defects, primarily pores and oxides, and that the presence and distribution of these defects are highly sensitive to the process conditions. Significant variations of defect distribution were, however, found in castings produced under the same conditions, indicating the stochastic nature of defects in die castings. The number of defects revealed on fracture surfaces varied between 2 and 20. The presence of porosity is mainly ascribed to gas entrapment phenomena during the die filling and to the blockage of vents due to the premature solidification of the molten metal. Generally, the gas pores showed a deformed spherical shape with shiny oxidized surface, while oxides appeared as rough and dull regions on the fracture surface. The presence of cold shots was also observed, generally, incorporated within porosity (Figure 3). Contrary, sound specimens revealed a fracture mode predominantly inter-granular with regions of cleavage facets, which are visible in the silicon precipitates and brittle inter-metallic phases, and with zones of deformed and fractured micro-necks of α-Al solid solution (Figure 4). The fracture path follows mainly the inter-dendritic eutectic zone. As previously observed by Cáceres and Selling, the fracture surfaces were flat, although a certain degree of tortuosity was present. Therefore, trying to determine exactly if a particular defect was intersected by the fracture plane or the locus of the actual intersection was a rather arbitrary exercise. The determination of area fraction covered by defects was particularly critical for oxides, which were sometimes sitting along the main axis of tensile specimens, as shown in Figure 5. Thus, it was assumed that all visible defects on the fracture surface lied on a single cross sectional plane and the projected area was considered for the calculation.