Curtin University of Technology - Department of Mechanical Engineering including Mechatronics - Staff - Dr Ian Davies - Publications

[55] I. J. Davies, S. Sieng, T. Ogasawara, and T. Ishikawa, “Influence of oxidation on the micro-mechanical properties of a 3-D woven SiC/SiC composite tested in air at 1100 ^oC”, Ceram. Trans., 192 pp. 175-186 (2006).

Abstract: A major issue limiting the widespread utilisation of ceramic matrix composites is the poor oxidation resistance of the fibre/matrix interface, which is often comprised of pyrolytic carbon or hexagonal boron nitride. Indeed, the fibre/matrix interface shear strength, t, is known to be highly sensitive to oxidation damage, in addition to playing an important role in the macroscopic mechanical behaviour of the composite. One potential method to increase the oxidation resistance of composite specimens has been to coat or seal the composite surface with an oxidation barrier material such as glass or pure silicon carbide (SiC). In the present work the authors have investigated the spatial distribution of micro-mechanical properties (specifically fibre strength and Weibull modulus, in addition to t) for the case of a partially oxidised 3-D woven composite based on the SiC/SiC system creep tested in air at 1100 ^oC. The composite was comprised of Tyranno^® LoxM Si-Ti-C-O fibres embedded within a matrix derived from the repeated polymer impregnation and pyrolysis of a precursor similar to polytitanocarbosilane. The composite had been coated (and any open porosity sealed) with a layer of glass following machining to the final test geometry. In addition to micro-mechanical properties, preliminary data was also presented for the spatial distribution of fibre pullout length within the composite fracture surface.

[54] I. J. Davies and R. D. Entwistle, “Influence of geometrical irregularities on the creep behaviour of ceramic fibres”, Ceram. Trans., 192 pp. 163-174 (2006).

Abstract: Due to the variation in fibre radius for current ceramic fibers based on the silicon carbide (SiC) system, a sufficiently long gauge length may span a region where the fibre radius varies by up to 50%. The potential effect of this variation on initial creep rate, e*, and creep rupture time, t_r , were investigated as functions of the radius variation, r_v (0 < r_v < 0.5), creep stress exponent, n ( 0 < n < 3), and the creep threshold stress, s_o ( s/s_o = 1.1, 1.5, 2.0). It was found that the initial creep rate, e*, increased compared to that of a prismatic fiber with the same mean radius as both n and r_v increased whereas the time to rupture decreased. The creep threshold stress was found to have little effect for the values of n encountered with current SiC-based fibres. It was concluded that researchers should be aware of the variation of fiber radius within their chosen gauge lengths and account for the effects noted in this paper.

[53] M. Sato, K. Itatani, T. Tanaka, I. J. Davies, and S. Koda, “Effect of chopped Si-Al-C fiber addition on the mechanical properties of silicon carbide composite”, J. Maters. Sci., 41(22) pp. 7466-7473 (2006).

Abstract: Silicon carbide (SiC) composites containing 0~50 mass% of chopped Tyranno^® Si-Al-C (SA) fiber (mean length: 214 mm (SA(214)), 394 mm (SA(394)), and 706 mm (SA(706)) were fabricated using the hot-pressing technique at 1800 ^oC for 30 min under a uniaxial pressure of 31 MPa in Ar atmosphere. The maximum flexural strength of the SiC composite was 344 MPa for 30 mass% of SA(706) fiber addition, whilst the maximum fracture toughness was 4.7 MPa·m^1/2 for 40 mass% of SA(706) fiber addition. Increasing the mean fiber length from 214 to 706 mm decreased the flexural strength from 380 to 281 MPa for 30 mass% of fiber addition, whilst the fracture toughness increased from 3.4 to 4.7 MPa·m^1/2 for 40 mass% of fiber addition. Through use of a treated SA(706) fiber containing an approximately 100 nm surface layer of carbon, the fracture toughness further increased to 6.0 MPa·m^1/2 for 40 mass% of fiber addition; this value was more than twice that of the monolithic SiC ceramic and is believed to be the highest so far achieved for this type of SiC/SiC composite containing chopped fibers.

[52] K. Itatani, A. Naitou, I. J. Davies, S. Sano, and S. Koda, “Formation process of magnesium aluminate due to solid-state reaction of highly-dispersed and nanometer-sized particles”, J. Soc. Inorg. Mater. Jpn., 13 pp. 336-344 (2006).

Abstract: The formation process of magnesium aluminate (MgAl₂O₄) due to the solid-state reaction of highly-dispersed and nanometer-sized aluminum and magnesium compounds has been examined by high-temperature X-ray diffractometry (HT-XRD), synchrotron radiation diffractometry (SRD) and X-ray photoelectron spectroscopy (XPS). The starting compounds were α- and γ-aluminum oxide (α- and γ-Al₂O₃; primary particle sizes, 105 and 31.6 nm, respectively) as aluminum sources, and magnesium oxide (MgO; 41.3 nm) and magnesium hydroxide (Mg(OH)₂; 61.1 nm) as magnesium sources. Through the combination of these compounds, four powder mixtures were prepared, namely, (i) α-Al₂O₃ and MgO, (ii) γ-Al₂O₃ and MgO, (iii) α-Al₂O₃ and Mg(OH)₂, and (iv) γ-Al₂O₃ and Mg(OH)₂. Phase change investigation during the heating of these mixtures indicated that the formation of MgAl₂O₄ due to the reaction of γ-Al₂O₃ with Mg(OH)₂ was faster when compared to the other combinations; almost single phase of MgAl₂O₄ could be obtained when this mixture was heated at 1200^oC for 1 h. More detailed investigation on the formation process of MgAl₂O₄ was conducted using the precursor mixture of γ-Al₂O₃ and Mg(OH)₂ heat-treated at 800^oC for 1 h. The data obtained from SRD and XPS suggested that small amounts of MgAl₂O₄ and α-Al₂O₃, together with γ-Al₂O₃ and MgO, were present in this precursor. The formation of MgAl₂O₄ due to the reaction of γ-Al₂O₃ with Mg(OH)₂ was found to occur readily due to active mass transfer as a result of the very small primary particle and agglomerate sizes.

[51] N. Tezuka, I. M. Low, I. J. Davies, M. Prior, and A. Studer, “In situ neutron diffraction investigation on the phase transformation sequence of kaolinite and halloysite to mullite”, Physica B, 385-386 pp. 555-557 (2006).

Abstract: Kaolin is a major raw material for the fabrication of conventional ceramics. In this work the authors have investigated the thermal phase transformation of mullite from two different types of kaolin (kaolinite and halloysite), with or without alumina matrix constraint, during heating up to 1500 °C and then cooling using in situ neutron diffraction. Mullitization was initiated upon heating to 1200 °C for all specimens and followed spinel formation at 1100 °C. Above this temperature, however, evolution of the main phases, i.e., mullite, cristobalite and corundum, was influenced by the presence of impurities, initial type of silica, and alumina constraint. The relative amount of mullite was largest for the pure kaolinite specimen, particularly during heating, and this was attributed to the presence of a glassy phase. However, kaolinite with alumina suppressed the crystallization of cristobalite from the glassy phase upon cooling due to a reaction between alumina and amorphous silica, consequently resulting in an amount of mullite as for the pure kaolinite specimen (approximately 65 wt%). Halloysite was less active in terms of mullitization due to the lower level of initial impurities and greater amount of cristobalite, particularly for the alumina-constrained specimen. However, the final amount of mullite derived from the pure halloysite specimen was similar to that as from the kaolinite specimen.