The characters of mold separation of cast bodies were investigated and discussed in relation to the mold separation coefficient, G, which has been reported in the previous paper of this series, V. The obtained conclusions are summarized as follows.
The mold separation coefficient, G, was influenced by three main factors, i.e. the permeability of water in cast pieces, the adhesive force in mold-cast interface and the shape of cast pieces involving their thickness. It was confirmed experimentally that G could be expressed as a function of the permeability, K(F₂/F₁-1)/l, by the equation G=a[logK(F₂/F₁-1)/l+b] where a was a constant depending upon the shape of cast pieces, and b could be determined by adhering nature at mold-cast interface. The permeability decreased markedly with increasing degree of deflocculation and thickness of cast pieces. Such mold-surface-treatments as forming a thin layer of fired powder of slip materials or sponging with water, which have often been utilized in factories, caused change in b values. The b values tended to be larger as an excess amount of deflocculant was added. The b values would be increased with increasing specific surface of slip materials.

View full abstract

Crystalline phases precipitated in glasses having the chemical composition of fluor-richterite containing aluminum-ion (Na⋅NaCa⋅Mg_5-xAl_x⋅Al_xSi_8-xO₂₂F₂, x=0-2.0) were studied by X-ray powder diffraction method, X-ray fluorescense spectroscopy and infrared absorption spectroscopy. The results obtained were summarized as follows.
1) Three crystalline phases were identified as fluor-amphibole-I, fluor-amphibole-II and fluor-phlogopite.
2) Fluor-amphibole-I precipitated in the glass where x=0.5, 1.0 or 1.5 had the structural unit of Na⋅NaCa⋅Mg_5-XAl_X⋅Al_XSi_8-XO₂₂F₂. The X-value, being very close to the x-value, increased with the x-value.
3) Fluor-amphibole-II precipitated in the glass where x=2.0 had the structural unit of Na_1-a⋅Na_1-bCa_1+c⋅Mg_5-XAl_X⋅Al_XSi_8-XO₂₂F₂. The X-value was smaller than the x-value. This composition was the intermediate one between fluor-richterite and fluor-tremorite.
4) Fluor-phlogopite precipitated in the glass where x=1.0, 1.5 or 2.0 had the structural unit of Na_1-aCa_0.5a⋅Mg₃⋅AlSi₃O₁₀F₂. The a-value varied with the x-value.
5) The coordination number of aluminum-ion in the crystalline phases were determined from the value of the chemical shift of the AlK_αX-ray emission line.
6) The relationship between the composition of fluor-amphibole-I and the wave number of infrared absorption peak associated with the Si-O symmetrical streching vibration was presented as the empirical formula.

View full abstract

This investigation was made to clarify how the indentation hardness of optical glass depends upon the effect of densification or plastic flow.
To examine the effect densification has upon indentation hardness, optical glasses with Knoop indentations were heat-treated at temperatures below their transition temperature. After heat-treatment, the percentage of change in the longer diagonal length of Knoop indentations was measured. These experiments were carried out on fused silica, BK, K, KF, F, SF, SK, SSK, BaF and BaSF glasses. Indentation recovery was found to be very small in optical glasses. Further, it was found that SF glasses which have the largest indentations show little or no change, while BK glasses which have small indentations show a much greater percentage of change. These results show that indentations in optical glasses are not always related to densification.
Using an interference microscope, Knoop indentation perimeters were observed in BK, KF, LLF, LF, F, SF, SK and SSK glasses. Without exception, a hump was found to be formed entirely around each indentation. This humping phenomenon shows that plastic flow occurs during indentation. Moreover, it was found that the ratio of the width (c′) of the hump around the shorter diagonal to the shorter diagonal length (a′) is approximately proportional to c/a calculated from Hill's theory on plastic-elastic materials, where (a) is the spherical radius of the cavity generated by a pressure and (c) is the spherical radius of the plastic region around the cavity. This means that Hill's theory can be applied to the indentation phenomenon of optical glass and accordingly the indentation of optical glass is always related to plastic flow.
The relationship between Young's modulus and glass composition was found to be almost parallel to the relationship between Knoop hardness and glass composition. Both Knoop hardness and Young's modulus increased with the content of net-work former. This supports Hill's theory. It was also found that c/a decreases with the content of glass net-work former.
From these experimental results, it would be concluded that the indentation hardness of optical glass, which contains a large amount of net-work modifier, is far more an expression of resistivity to plastic flow than densification.

View full abstract

Employing the author's method for calculating the packed volume of multi-component particle systems, the effects of particle-size distribution on packing density of silicon carbide bodies, packed by moulding method with stick penetration, were calculated.
From the results, it was predicted that the packing density would increase with the number of particle components involved and, in the case of gap grading, a better packing could be obtained when a component close to the coarsest fraction was removed from the system. Another prediction was that a closer packing could be obtained when the value of q, the Andreasen continuous-grading coefficient for particle below 1.41mm, fell between 0.5 and 0.8.
In the case of packing by press moulding, however, it was surmised that a particle-size distribution with still lower value of q would be more adequate for a closer packing.
The results of packing experiments by moulding method with stick penetration shown in the another report were in good agreement with tendency for those of calculations. Successive experiment by press moulding showed that, a closer packing was obtained when q fell between 0.5 and 0.6, and also when the mix was devoid of a component close to the coarsest fraction. These results agreed well with former predictions.
A mix containing a large amount of coaser fractions had a tendency to approach to the particle-size distribution corresponding to the closest packing, because of crushing of grains in pressing.
It was also found that weight increases of a pressed body in firing, firing expansion and hot strength of a fired body were all dependent on the q value, in that they increased as q decreased, and their correlation with q was closer than with the mean particle diameter.

View full abstract

Densification process of the bodies consisting of kaolin and quartz grains was investigated kinematically by means of isothermal shrinkage measurement, and microstructurally by X-ray diffraction method and observation with polarization and scanning electronmicroscopes.
From the results of linear expansion and shrinkage measurements of the specimens, the bodies showed distinct shrinkage of three stages occurred at about 500°, 900° and 1100°C or higher temperatures.
It was inferred that the shrinkage occurred at about 500°C was caused by decomposition of kaolin minerals, and the shrinkage decreased with increase of quartz content of the body, because of the expansion associated with α-β transformation of quartz at 573°C. The shrinkage appeared at about 900°C seemed to occure with diffusion sintering process at metakaolin grain boundary. Apparent activation energy of the shrinkage process was increased with increase of quartz content of the body. The increase might be caused by interruption to the grain boundary diffusion with presence of quartz grains incorporated in the body. The rate constant of the shrinkage at 900°C was also decrease with increase of the quartz content.
The shrinkage observed at about 1100°C might be associated with densification process occurred by viscous flow of spinel type intermediate phase grains with small amount of melt formed from a little impurities contained in kaolin parts. Apparent activation energy and frequency factors for the shrinkage process were not affected by the presence of quartz grains, but addition of the small amounts of quartz tended to promote densification of the bodies, and further addition of quartz decreased markedly rate constant of the shrinkage at 1060°C.
The bodies shrank more rapidly at a higher temperature than 1100°C. The shrinkage also seemed to be due to viscous flow, but it diminished when almost all the pores in the body were closed in the glassy matrix. Further, changes of the constituent minerals and microstructure of the bodies with the densification process were described.

View full abstract

Batches of photochromic glasses (about 100g) containing various amounts of Ag and Cl (the nominal oxide compositions, SiO₂ 52.63, B₂O₃ 20.09, Al₂O₃ 6.89, Li₂O 2.90, Na₂O 1.80, BaO 8.18, PbO 4.81, ZrO₂ 2.10, Cu₂O 0.014, Ag 0.44-5.80, Cl 0.15-2.18wt%) were melted at 1400° or 1450°C in the air or in the air saturated with AgCl, NaCl or BaCl₂ vapor and amounts of Ag and Cl retained in the glasses after melting for various times were measured by X-ray fluorescence analysis.
1) The glass batches containing large amounts of AgCl (more than about 3wt%), when melted in air, first separated into two liquid phases; a phase rich in AgCl and that rich in silicate. The latter remained as glass after cooling to room temperature. A great part of the AgCl added to the batches was thus lost almost instantaneously from the glass. After the phase separation, losses of the Ag and Cl from the glass melts continued probably by evaporation of chlorides such as AgCl and NaCl from the melts and also by oxidation of the chlorides in the melts by air: 4AgCl+O₂→2Ag₂O+2Cl₂↑ In the latter mechanism only the Cl was considered to be lost because the Ag₂O formed has high solubility in the glass melts. Loss rates of the Ag and Cl were both low but that of the Cl was a little higher than that of the Ag.
2) The glass batch containing small amounts of AgCl (less than about 0.5wt%) showed no liquid phase separation but losses of the Ag and Cl occurred at slow rates as observed for the batches containing large amounts of AgCl.
3) When the glasses were melted in the air saturated with AgCl vapor, the Ag contents increased rapidly, whereas the Cl contents were held constant. When the glasses were melted in the air saturated with NaCl vapor, the Ag contents decreased at slow rates, whereas the Cl contents were held constant. The loss rate of the Ag in the air saturated with NaCl vapor was a little higher than that in air. Effects of the air atmosphere saturated with BaCl₂ vapor were intermediate between those in the air and the air saturated with the NaCl vapor.
The above results were explained by supposing suppressing effects of the chloride vapors in the atmosphere on evaporation of the chlorides from the molten glasses and oxidation of the chloride vapors by air.
4) Mechanical stirring of the molten glass did not give remarkable effects in accelerating evaporation of the chlorides from the molten glasses.

View full abstract