Process optimisation in the semi-solid forming of hypereutectic Al/Si MMCs

A novel processing route for the fabrication of Allhigh silicon MMCs is presented. The silicon size is much finer than can be achieved by casting, yet the materials can still be formed into a near-net shape. Initial properties of the MMCs are presented, and methods under investigation to optimise processing and improve properties are discussed. fntroduction Hypereutectic AVSi ANoys Hypereutectic AIlSi alloys have been used for some years in the automotive industry, and in other fields where wear resistance, high specific stiffness and thermal stability are required (1). Typical uses for hypereutectic AIISi alloys are: pistons, engine blocks, connecting rods and pump and hydraulic components. Such components are usually made using conventional casting techniques, or by DC casting, hot working and machining (2). The most commonly used alloy is the 390 series (3), which contains 16-18 wt. % silicon, as well as 5% copper and 0.5% magnesium. This alloy possesses a higher wear resistance than most aluminium alloys, and 10-15% greater specific stiffness, along with -15% lower coefficient of thermal expansion. Although such improved properties are significant, increased silicon loading holds out the promise of even greater improvements, moving into the realm of a true composite, if certain processing difficulties can be overcome. Hypereutectic AIISi alloys possess two advantages compared to other AI-MMCs: the particulate phase is formed in situ from the melt, so there is no reaction at the AIISi interface and particulate-matrix bonding is very strong forming from the melt also means that the MMCs can readily be recycled, or used as components in alloys lower in silicon. This will have a significant effect on the cost of the alloys, as refining Al consumes 20 times as much energy as remelting it (4) The source of the processing difficulties is the steep liquidus in the hypereutectic region of the AI/Si phase system. The solidification range of the alloys rises rapidly with silicon content, with the consequence that the primary silicon phase soon becomes coarse, even when phosphorus refinement during casting is used. Refinement has been shown in the present study to have an appreciable effect on silicon size and morphology at Si contents up to 36wt. %, reducing the maximum particle size achieved at moderate cooling rates from 300pm to 100vm. However, Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jp4:19937268 1712 JOURNAL DE PHYSIQUE IV this is still too large, and properties such as strength, toughness and ductility are compromised. In addition, at these high Si contents, it is necessary to perform the melting and refinement under an inert atmosphere. This itself is a consequence of the high melt temperatures necessary for casting in the hypereutectic region, as it appears that the aluminium phosphide nucleant oxidises at temperatures in excess of 750°C. (A1136Si was cast into a copper wedge mould at 900°C). The silicon content of the 390 series of alloys corresponds to the maximum casting fluidity. It has been found that the wear resistance of cast hypereutectic AIISi alloys is not improved significantly by Si loadings above 16-18% (3), so it is reasonable to choose a composition which facilitates processing. This is especially the case when hypereutectic AIISi alloys are found to possess rather poor shrinkage feeding characteristics, requiring high melt superheat and mould temperature gradients and complex multiple gating systems in order to promote directional solidification and a uniform Si distribution (3,5,). The steep liquidus in hypereutectic AIISi alloys also means that practical problems become greater at higher Si contents. Thus, in order to achieve a loading of 25% Si, a melt temperature of -800°C is required, with the attendant difficulties of wear of dies and refractory components, oxidation and gas pick-up in the melt (5,6). Spray-forming and Thixoforming Hypereutectic AI/Si MMCs The approach of the present project (which is a BRITEIEURAM collaboration, involving six Partners, from Spain, Britain and Germany), is to try to circumvent some of the problems encountered with casting hypereutectic AIISi alloys, and thus to achieve higher silicon loadings, while maintaining a fine silicon particle size. This can only be done by using a rapid solidification technique. In summary, the route chosen was as follows: Production of preforms using the Osprey process (71, in which the melt is atomised in an inert gas and collected on a rotating former. The resulting preform possesses a non-dendritic microstructure suitable for semi-solid processing (8). Spray-forming achieves the cooling rates typical of powder metallurgy (lo4-106k-'), while avoiding some of the processing steps Machining of the preform and extrusion into a round bar. This eliminates any residual porosity from the spray-forming, and in the case of the hypereutectic AIISi system, breaks up the continuous silicon network (9). Near-net shaping of billets machined from preforms or cut from the extruded bar by thixoforming. The billet is heated into the semi-solid region, where, because it is non-dendritic, it behaves in a thixotropic manner. This means that when left undisturbed the billet is solid, but when sheared it flows like a liquid. Thixoforming restricts microstructural coarsening as the processing temperatures and liquid contents are low. Alongside this advantage, which is of particular importance for the hypereutectic AIISi system, thixoforming offers a number of other benefits. The low temperatures and forming pressures reduce solidification shrinkage and die attrition, and allow the use of die materials which are more easy to machine than tool steel. Laminar flow of the slurry in the die reduces gas and lubricant entrapment, and also particle segregation when composites are thixoformed. Solidification occurs under pressure: when suitable die designs are employed, shrinkage porosity can be eliminated.