Microstructure and mechanical performance of SiC/BMAS glass-ceramic matrix composite
A diverse range of microscopy techniques and mechanical testing methods have been used to characterize glass and glass-ceramic composites. The focus of the work has been a commercially available Barium Magnesium Aluminosilicate matrix reinforced by Tyranno SiC based yam type fibres. The mechanical behaviour has been related to the microstructure through use of models from the literature. The temperature range of study has been from room temperature to 1300°C in air. The microstructure of the BMAS(fyranno was a diphasic mixture of celsian and indialite/cordierite although the manufacturers intention was a monophasic bariumosumilite. The carbon rich interface was found to be thin (l0-15nm) but the composite displayed impressive strength when compared to similar glass-ceramic composites reported in the literature. The matrix could be converted to the equilibrium bariumosumilite phase by heating in an inert atmosphere at 1370°C (or possibly lower) but matrix elemental diffusion into the fibres is likely to impair fibre strength. Tensile failure was by conventional matrix microcracking with load transfer to the in line fibres. However the composite strength was found to be dependent upon the strain rate as was the microcracking threshold associated with cracking of the 0° plies. Failure of the UD BMAS(fyranno was by longitudinal splitting before the expected ultimate strength (from the 0,90° results) was reached. This was due to an apparent notch sensitivity in this fibre architecture, a trait not observed in the 2-D composite. Direct measurement methods were used to establish the interfacial shear strength and these were compared to various models. These were based on matrix cracking thresholds, matrix crack spacing and a relatively new method where an 'inelastic strain index' was found from loading and unloading curves or hysteresis loop widths. Greatest fidelity with the direct methods was found with the last of these models. As with all composites with carbon enriched interfaces oxidation of the interface and fibres was found to impair strength when tested in air at temperatures as low as 600°C and possibly below this when testing at lower strain rates. At high strain rates, near room-temperature-strengths were achieved, even at 1l00°C, as the degrading effects of the oxidizing environment had less time to act. Long term exposure at high temperatures (1200°C) was responsible for formation of an embrittled surface layer up to 70J.lm thick. Within this layer the fibres were severely degraded and strong bonding prevailed at the interface. At temperatures in excess of the expected fibre pyrolysis temperature, (l100°C), the composite was seen to shrink along the length of the fibre axis and dilate normal to it which was attributed to fibre instability. Stabilising the fibres by heat treatments at 1200°C for 24 hours was seen to improve the creep performance in terms of the total strain accumulated within the 100 hours of the creep tests. The creep was comparable to other commercial glass ceramics (CAS/Nicalon and BMAS/BN/SiC/Nicalon) indicating the dominance of fibre creep properties on those of the composite. Cycling of the creep load seemed to result in a greater embrittled depth from the surface but failure at 100MPa and 1200°C was not observed within 240 hours of testing. Other systems were investigated such as the CAS/Nicalon, MAS/Nicalon and AS/Nicalon. Of these the AS/Nicalon was used for modelling the creep behaviour since it represented a simple system where matrix creep was accompanied by elastic deformation of the fibres. A model from the literature was used to explain an apparent increase in the elastic modulus during load cycling at high temperature and also the lower strain accumulation seen during load cycling compared to conventional creep tests.