| Summary: | A survey of the current literature on binary eutectic systems shows that, although the simpler eutectic morphologies produced between phases of low entropy of fusion are now fairly well documented and understood, the same is not true when one of the phases has a high entropy of fusion. This thesis is an attempt to partially rectify this dearth of fundamental information on metal/non-metal eutectic systems. The system aluminium / silicon was chosen as being one typical of this class of eutectic alloys, High purity specimens were undirectionally solidified under carefully controlled and treasured growth conditions. The resultant microstructures were then classified in terms of the growth variables of</p> <p align=center>R - the growth rate</p> <p align=center>G - the temperature gradient in the liquid ahead of the growing interface</p> <p align=center>C - the alloy composition The growth variables were found to greatly affect the mode of the eutectic reaction and to cause not only changes in the scale of the eutectic microstructure but to alter the form of the microstructure as well. The undirectionally grown specimens were subjected to detailed morphological and crystallographic analyses and in this way information was obtained about the possible growth mechanisms of these various forms of the eutectic microstructure. The various eutectic forms were as follows:</p> <p align=center>a) Long range Diffusion When the Al/Si specimens were undirectionally solidified above a certain critical G/R value (G/R = 10<sup>7</sup>sec.°C.cm<sup>-2</sup>), the two eutectic phases grew by means of a relatively long range diffusion process. The aluminium phase grew with an isothermal interface and the silicon grew with a non-isothermal interface, to produce a massive form of the eutectic. The interconnected silicon phase grew by a twin plane reentrant edge growth mechanism.</p> <p align=center>b) Short range Diffusion Below this critical G/R value, the two eutectic phases grew by a short range diffusion process, which could be either steady or non-steady state depending on the growth conditions. The form of these short range diffusion types depended upon the growth kinetics of the silicon phase and the ease of aluminium overgrowth.</p> <p align=center>i) <100> Si growth textured form At low growth rates (R ~ 10<sup>-5</sup>cm/sec.) the silicon phase grew by means of a screw dislocation kinetic mechanism and this produced interconnected silicon crystals elongated in the growth direction which had a <100> Si growth texture; the width of these particles being affected by both growth rate and temperature gradient. This form of the eutectic silicon normally grew by a steady state and short range diffusion process. Corrugated silicon particles were observed, however, indicating that a non-steady state growth mechanism could also operate.</p> <p align=center>ii) {111} Spinel twin related form When the pure eutectic alloy was unidirectionally grown at higher growth rates (R > 3 x 10<sup>-4</sup>cm/sec.) the eutectic silicon phase grew by a different kinetic mechanism which involved the use of {111} spinel twins. Growth was normally by a non-steady state process and this produced an irregular microstructure. The eutectic silicon in this form of the alloy was always found to be interconnected, even at the highest attainable growth rates, (i.e. rapid quenches) and it was shown that isany of the irregular silicon particles grew by means of a twin plane reentrant edge mechanism and that the various particles were related to each other via multiple {111}; spinel twins. In hypereutectic alloys this same T.P.R.E. growth mechanism for the silicon phase produced a steady state complex regular structure with a <100> Si growth texture in addition to the irregular form already mentioned above.</p> <p align=center>c) Repeated Nucleantion and Growth No experimental evidence was obtained to show that Al/Si eutectic alloys ever grew by a process of repeated nucleation and growth. The eutectic silicon phase was always found to be interconnected.</p> <p align=center>------------------------------ Unidirectional solidification experiments with the other F.C.C. metal/silicon eutectic systems (Ag/Si and Au/Si) produced microstructures which showed the same silicon morphology and crystallography as the Al/Si system. Finally, in the light of the work on the pure Al/Si eutectic system, a series of controlled experiments were carried out on the modification of the alloy by metallic sodium. Results showed that the modified eutectic silicon was fibrous and interconnected and very similar to the rapidly quenched pure eutectic. It was concluded that the addition of sodium to Al/Si alloy must produce the change in microstructure by affecting the growth and not the nucleation of the eutectic silicon. From the experimental results it appeared that the sodium was affecting the growth of the eutectic silicon by preferential poisoning of twin plane reentrant growth sites.
|
|---|
