Half-maximum line breadths were measured on main reflections of synthetic 11Å tobermorite. K_α doublet separation and instrumental line broadening were corrected on them.
Firstly, separations of crystallite size and lattice strain were tried on the corrected half-maximum line breadths using the equation of Williamson et al. The results showed lattice strain was negligible for 11Å tobermorite. Therefore, crystallite sizes, i.e. D₀₀₂, D₂₂₀, D₄₀₀ and D₀₄₀, were calculated from the corrected half-maximum line breadths of 002, 220, 400 and 040 reflections respectively by the Scherrer equation assuming the shape factor 0.9.
D₀₀₂ corresponded to the crystal thickness of 11Å tobermorite calculated from the BET specific surface area neglecting edge effects.
A volume of a single crystallite was calculated assuming that the shape of crystallite was disk-like of D₂₂₀ in diameter and D₀₀₂ in thickness. A good linearity was observed between the calculated crystallite volume and the crystallinity. This relationship also suggests that a main factor to determine the crystallization process is not an increase of number of crystallites, but a growth of crystallite size.
Isomorphous substitution of Al ion for Si ion in the lattice of 11Å tobermorite reduced D₀₄₀. This is thought to be caused by disorder resulted from the substitution of the bigger Al ion in ion radius for the smaller Si ion.

View full abstract

Solubility behaviors of glass fiber for water by making under various spinning conditions were studied.
Solubility of Na₂O and SiO₂ from glass fibers were increased with the decrease of the spinning temperature and with the increase of the spinning velocity.
Especially, solubility of SiO₂ from glass fiber was markedly higher than that of the bulk glass and the annealing glass fiber.
It was inferred from these results that surface structure of glass fiber was modified by various thermal history.

View full abstract

The solid solubilities in the alkaline-earth metal aluminates with the stuffed tridymite structure (CaAl₂O₄, SrAl₂O₄ and BaAl₂O₄) were investigated on heating at 900°-1500°C in air. The results obtained are summarized as follows:
(1) The soild solubilities in these aluminates, CaAl₂O₄, SrAl₂O₄ and BaAl₂O₄, depend on the difference in the linkage of AlO₄-tetrahedra in the stuffed tridymite structure. Two types of solid solution, BaAl₂O₄-type and CaAl₂O₄-type, formed in the binary and the ternary systems in CaAl₂O₄-SrAl₂O₄-BaAl₂O₄.
(2) The BaAl₂O₄-type and the CaAl₂O₄-type structures are closely related to the average ionic radius of alkaline-earth metals which were stuffed in the channels made up of AlO₄-tetrahedra. In the range of 1.14 to 1.21Å, the CaAl₂O₄-type structure was formed, and from 1.23 to 1.50Å the BaAl₂O₄-type was formed.
(3) Below 1.37Å of the average ionic radius of alkaline-earth metals, the BaAl₂O₄-type structure slightly distored to the α-SrAl₂O₄-type structure at low temperatures.

View full abstract

Studies on the electric conduction of the glass in the system CuO-V₂O₅-P₂O₅ have been carried out to elucidate the electronic conduction by the hopping between two kinds of transition metal ions and the switching phenomena of highly conductive glass.
The composition of the glass containing maximum amounts of transition metal oxides was found to be 0.25-0.4CuO, 0.7-0.55V₂O₅ and 0.05P₂O₅. The existence of low valence oxides was confirmed by the weight increase as heating the glass in oxidation atmosphere.
The electric conductivity of the 4V₂O₅⋅P₂O₅ glass at 200°C increased from 2×10^-3 to 7×10^-3Ω^-1cm^-1 by the addition of CuO from 0 to 40mol%, which decreased the amount of V₂O₅ from 80 to 50mol%. The conductivity becomes about one hundredth by the corresponding decrease of V₂O₅ in V₂O₅-P₂O₅ glass. Therefore, CuO seemed to contribute to the electric conduction by the same magnitude as V₂O₅.
The abrupt increase of conductivity, that is switching, was observed when the electric field higher than the threshold voltage V_th was applied between the wedge electrodes on the surface of the glass. The relation between logV_th and logR (the resistivity of glass between electrodes) was nearly linear havinng the inclination of 2. LogL (inter-electrode distance) was also linear to logE_th(V_th/L) and its inclination was -0.85. The relation between the threshold switching and the Joule heat was considered from these results. However, from the measurement of the increase of electric current with applying time of various high voltage, it was found that the Joule heat given by electric current before switching started became rather smaller, the higher the applied electric field. Therefore, the switching seemed to be caused not only by the Joule heat, but also by some other effects, such as the increase of conduction electron by applying high electric field.
In the cage of applying the field higher than the threshold voltage, the dendrite crystalline being rich in Cu oxide developed in the glass from cathode to anode. The deposition of crystalline was attributed to the migration of Cu ion. The more drastic change of the glass was the formation of the molten path between electrodes. The threshold voltage decreased at the second cycle, because the conductivity of the developed crystal or the molten path was higher than the original glass.

View full abstract

A three dimensional finite element analysis was used to calculate thermal stresses caused by temperature gradient in refractory heated from one end. Excellent agreement was obtained between calculated and experimental results.
Although thermal stresses produced in refractory increase with increasing temperature, comparative stress distribution remains almost unchanged. That is, the hot face region and the side region are in compression and these are backed by a region in tension around the center axis with a peak at about 50mm from the hot face. Following tensile stresses are considered to cause cracking.
(i) Normal stress component perpendicular to the hot face around the center axis at about 50mm from the hot face.
(ii) Normal stress component parallel to the hot face around the center axis at about 30mm from the hot face.
As the specimen dimension increases, thermal stresses increase with the maximum tensile region moving away from the hot face. The magnitude of the normal stress component perpendicular to the hot face is scarcely affected by the restraining stress in the direction parallel to the hot face. On the other hand, normal stress component parallel to the hot face almost results in the superposition of the thermal stress component and the restraining stress.

View full abstract

The compacts of the mixture of siliceous sand and Al were heated in N₂ atmosphere in a presence of carbon material on the route of gas flow. The weight change of the compacts was measured by thermogravimetric apparatus from a room temperature to 1500°C, and the phases in the compacts at various stages of heating were analyzed by using X-ray diffraction technique.
Nitriding of Al took place above about 500°C and its reaction rate increased at the melting point of Al (660°C), The exothermic reaction of AlN formation induced the reaction between siliceous sand and Al to form Si and α-Al₂O₃. AlN⋅Al₂O₃ was formed in the compacts heated at the temperature range of 660° to 1200°C. Si reacted with N₂ to form β′-Si₃N₄ above 1100°C and 15R-AlN was also formed in the compacts heated above 1300°C. Silica rich phases such as mullite, O′ and X phases were observed in the compacts which did not contain enough Al to reduce silica in the siliceous sand.
With increasing temperature, weight loss was observed and became considerable above 1400°C. From the phase change in the compacts, the weight loss was considered to be mainly caused by the evaporation of SiO formed through the following reaction:
Si₃N₄+Al₂O₃=3SiO+2AlN+N₂
Mullite, O′ phase and X phase disappeared from the compacts when they were heated at 1500°C for about 27h. They were considered to be reduced by CO gas which was formed by the reaction between carbon material and oxygen present in N₂ as impurity.
The weight loss kinetics at 1500°C followed parabolic rate law. The pore structure of the compacts seemed to have effect on the permeability of the gas phases.

View full abstract

Composition of the mullite like compound (often called as X-phase in Si-Al-O-N system) were studied by the following experiments.
(1) Correspondence between the composition of raw mixture and formed compound after hot-pressing the mixture at 1700°-1820°C.
(2) Analytical results of the X-phase with E. P. M. A. and wet chemical analysis.
(3) Correspondence between X-ray diffraction data collected with a “single” crystal of the X-phase and the analytical results.
From the results of those experiments, it was concluded that the X-phase has very close composition to that proposed by L. J. Gauckler et al., Si₄Al₄O₁₁N₂. The atomic ratio, Al/Si was considered to be variable in a range, 1.0-1.8, but the ratio N/O was considered to be in fairly narrow range, 0.17-0.18 at about 1700°C. Typical compositional formula, Si₆Al₇O₁₈N₃ was proposed for the X-phase under fully occupied state.

View full abstract

The formation of brookite, metastable phase of TiO₂, by the oxidation of titanium metal only in the presence of NaF solution at 200°-500°C, 10-400kg/cm² pressure, took place at the surface of Ti metal. On the other hand, anatase crystallized in most of solutions other than NaF. The brookite→rutile transformation was accelerated by increased temperature, reaction time, and solution acidity. The brookite crystals obtained were generally small (<5μm) and pale blue. Favorable conditions were found to be: crystallization temperature, 200°-300°C; reaction time, ‹1 day; and solvent, › 0.5mol/kg NaF.

View full abstract

In order to elucidate the mechanism of mass transport on glass surface, the amplitude decay was measured of multiple parallel surface scratches on lead silicate and borosilicate glasses in dry N₂ at 547.5°-610.7°C and 694.6°-714.8°C, respectively. The surface profile, which was close to square wave initialy, developed into a sinusoid by pre-annealing, and further annealing resulted in exponential decay of the amplitude of the sinusoidal wave with time. The decay rate constant was inversely proportional to the wave length of the profile. This finding led to the conclusion that the mass transport process occurred via the viscous flow mechanism. The viscosity calculated from the decay rate constant was larger than that obtained by the conventional fibre elongation method, indicating that the viscosity in the surface region was larger than in the bulk.

View full abstract