This paper deals with the stress which arose in outer glass bulbs of mercury discharge lamps during lighting. At their uses bulbs were irradiated by ultra-violet ray from inner discharge tudes made of fused silica and cooling water flowed around outside the bulbs.
Two kinds of glass bulbs, namely of Pyrex type borosilicate and 96% silica glasses, were examined.
Severe tension stresses concentrated at the inner surface were observed photo-elastically, and compaction (density increase) of glass in the same surface layer was also recognized. By heating up, the stress began to release itself at comparatively low temperatures; for borosilicate glass about 250°C and for 96% silica glass about 600°C. By irradiation of a glass test piece with a mercury discharge lamp, experimental reproduction of concentrated surface tension was possible.
Although the stress seems to result from compaction of glass by ultra-violet irradiation, photo-chemical, structural or atomic interpretations are not given now.

View full abstract

This study was made of the relationship between glass stones in brown bottle glass and impure minerals in quartz sand and dolomite which were used as raw materials.
From the mineralogical study of the stones, the bulk of them was found to consist of silicious and aluminous minerals which were introduced from quartz sand and refractory bricks. Some amounts of spinels and other optical isotropic minerals were found as constituent minerals of stones.
As impurities in raw materials following minerals and rocks were found;
Quartz sand: magnetite, hematite, ilmenite, chart and hornfels.
Dolomite: diopside, hornblende, mica, anorthite, glossularite, spinel, graphite, magnetite, pyrrhotite and lamprophyer.
Solubilities of the minerals in molten glass indicate that some amounts of spinel, ilmenite and glossularite probably remained as stones during glass melting process, and spinels which were found in glass stones had the same forms as in the dolomite. Chart and hornfels, however, cannot remain as insoluble silicious minerals because of their high solubilities.

View full abstract

Sintering of alumina is affected by the property of raw material, additives, atmosphere, etc. In this paper, the rate and the activation energy of sintering of alumina containing MgO, Co₃O₄, TiO₂ and MnO₂ by 2 or 5 wt.% were examined in the temperature range from 1400 to 1650°C.
The relation between the extent of sintering (P) and time (t) can be described by the following equation: P=Ktⁿ where K is the rate constant for sintering and n is the time exponent depending on temperature and specimens with different additives. A recent work on sintering of alumina has reported that n was a constant of the value 0.4 unrelated to temperature. However, n from this experiment was not constant but small at low temperatures and with rising temperature increased up to a maximum point and then decreased. The value was smaller than 0.4, and 0.2-0.29 at the maximum point though varied by the additives. The velues of n obtained were as follows:
The self-diffusion coefficient was calculated from the measurement of shrinkage using a Kingery's equation for diffusion sintering, from which the activation energy (Q) for sintering was obtained by Arrhenius plotting. The results were as follows:

View full abstract

The ceramic property of five kinds of Fukui pottery stones occured in Onigatake, Jogatani, Myohoji and Uesaka, was investigated.
The mineralogical composition was identified by means of X-ray diffraction, D. T. A., chemical analysis and electron micrograph methods.
The following properties were measured on 20hrs wet ball-milled samples, i.e. viscosity, plasticity, modulus of rupture, refractoriness and thermal expansion.
Pressed specimens of the wet ball-milled sample were fired in the temperature range of 800 to 1300°C and determinations of firing shrinkage, thermal expansion, water absorption, reflection, modulus of rupture and moisture expansion were made on the fired specimens.

View full abstract

Previously we reported that the grain growth rate of alumina during hot-pressing was slower than that during ordinary sintering but increase in temperatures during pressing caused very remarkable growth of alumina grains.
In the present experiment Linde-C alumina was hot-pressed in various ranges of temperatures, and the effect on grain growth phenomenon of alumina was examined. Alumina specimens were heated inductively up to 1800°C in 30 min. and cooled immediately in graphite molds. The pressure of 200kg/cm² was applied in various ranges of temperatures i.e. from room temperature to 1500°, 1600° or 1700 or from 1500°, 1600° or 1700° to 1800° or from 1600° to 1700°C.
Exaggeratedly grown grains (100 to 1500 microns in length) were observed in alumina heated at increasing temperatures in the range from 1600° to 1700°C under a constant pressure and completely densified. On the other hand, alumina heated up to 1600° or from 1700° to 1800°C under the same pressure consisted of fine grains and the average grain diameter at their central zones was smaller than 10 microns. These phenomena, were discussed with respect to strain energy and its relaxation; the strain energy was given by the plastic deformation of alumina grains during hot-pressing and was relaxed by recrystallization and subsequent grain growth.
According to M. L. Kronberg (J. Am. Ceram. Soc., 45, 274 (1962)) the yield stress of alumina single crystal is 220kg/cm² at 1600°C and at the strain rate of 0.002 in/in/min. So the applied pressure of 200kg/cm² slightly above 1600°C might give a slow and successive plastic deformation and strain to alumina grains. Such strain energy would make only few grains grow exaggeratedly, because it is two orders of magnitude larger than free energy at grain boundaries which is the driving force of ordinary grain growth.
The reason why alumina pressed above 1700° or below 1600°C did not show exaggerated grain growth were also discussed in respect of the amount of strain energy, the number of recrystallization nuclei and relaxation by recovery.
Alumina hot-pressed almost to theoretical density and then heated one hundred degrees additionally in several min. had a layer of coarse grains (30 to 100 microns in diameter) at their surfaces. Such grains were explained by the enhancement of normal grain growth by the temperature gradient within the specimen.
If ΔT is the temperature difference across the grain boundary and ΔF is the change in free energy on going across the boundary, the rate of the boundary movement is described in the next equation,
G=RT/Nh⋅l⋅exp(-ΔF*/RT)⋅ΔF*/RT(ΔT/T+ΔF/ΔF*)
where ΔF* is the activation energy of grain growth and l is the average distance of each atomic jump.
Comparing with the equation derived by N. F. Mott (Proc. Phys. Soc., 60, 391 (1948)) for the case with no temperature gradient,
G=RT/Nh⋅l⋅exp(-ΔF*/RT)⋅ΔF*/RT
the author concluded that the rate of the grain boundary movement was enhanced by the temperature gradient at the ratio of (ΔT/T+ΔF/ΔF*) to ΔF/ΔF*. When ΔT is assumed to be 0.05°C at the surface of the specimen, which is the case of the present experiment, the ratio becomes 5:1. This means that the rate of the boundary movement becomes 5 times faster than in the case of uniform heating.

View full abstract

This study was undertaken to determine the range of chemical composition of frits for the use of porcelain enamel on Aluminum plate, and crystaline phases which have been developed in the porcelain enamel.
It has been found that the oxygen potential of about 1.1 calculated from the chemical composition of frits, is most suitable in order to produce a good enamel for Aluminum plate, which is in agreement with the result in a previous paper (PbO-Sb₂O₃ system).
Crystals developed in the porcelain enamel prepared from frits in this system, have the same structure as perovskite or corundum. Both of crystals are considered to be composed of PbO and TiO₂, but in the latter BaO, CdO and V₂O₅ are likely to contribute to form the hexagonal crystal structure.
Attempts to prepare these crystals by firing the mixture of the oxide components were still unsuccessful.

View full abstract