The effect of alloy and impurity variation on the treatment, casting and physical properties of aluminium-silicon eutectic alloys
Thesis DisciplineMechanical Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
The aluminium-silicon eutectic alloy finds widespread use in commercial light alloy foundries world-wide. The intrinsic characteristics which ensure that this traditional alloy continues in use are excellent fluidity, moderate strength, good ductility, low shrinkage, no requirement for post casting heat treatment and exceptional corrosion resistance. The latter two are particularly noteworthy as they assure the eutectic alloy finds favour in numerous roles for which the higher strength Al-7%Si-Mg alloy is less well suited. The research reported in this thesis aimed to quantify the changes in physical and structural properties of sand-cast Al-Si eutectic alloy due to compositional variations within the specified range provided by British Standard 1490-LM6. The eutectic alloy was selected for investigation primarily due to members of the local (NZ) aluminium industry expressing concerns regarding consistent production of castings capable of surpassing the physical requirements of the BS1490-LM6 standard. To achieve the desired aim approximately 500 standard test bars (as used in industry and specified by BS1490) were produced using conditions replicating those encountered in a small, commercial foundry. Each specimen cast was of varying composition with the major variables being Na, Sr, Ti, B, Si, Mg, Mn and Fe. Physical and structural properties including: hardness, tensile strength, ductility, 0.2% proof strength, porosity, grain size and eutectic silicon morphology were monitored for each specimen produced. The combined composition and physical/structural data were then subjected to extensive statistical analysis via multi-linear-regression. The results of the statistical analysis are presented as a series of expressions relating the measured properties to the relevant compositional variables. The sometimes complex and inter-related effects of the elements responsible for significant property variation are illustrated in both numerical and graphical forms. The full analysis results and associated findings are too numerous to be summarised here. An example of a significant finding is that, with the exception of grain refinement, boron is detrimental to every property monitored. Almost without exception the unwanted effects of boron are already manifest at boron levels as low as 0.01%. Another finding of interest is that within the confines of the given compositional range, the controversial practice of adding manganese to counteract iron-induced embrittlement has little effect on ductility while being highly deleterious to tensile strength. Indeed the embrittling effects of iron were found to be far smaller in magnitude than anticipated given the concern which surrounds this element. This led to the conclusion that further investigation into the effects of iron on parameters such as fatigue endurance warrant investigation. Two compromises were unavoidable in this research firstly, an inability to assess the level and hence effects of phosphorus variations, and secondly the boron additive used was not of the desired type. Surprisingly, the effects of boron mentioned above were not found to be discernibly influenced by the latter compromise. Two less significant areas of experimentation were also touched upon in the course of this research, namely the suitability of various materials for use in conditions requiring contact with molten aluminium, and the contamination and compositional variations which occur during degassing by conventional lance and tablet methods.