From information furnished by the results of experiments and the mathematical analysis based on the “Glassymicro-phase theory”, the viscosity and visco-elastic properties of glass in transformation region the authors arrived at the following conclusions.
(i) The viscosity value of a glass obtained by applying constant load at a constant temperature was able to be separated into two domains by author's t_s curve (J. Soc. Glass Tech., 41, 137, 295 (1957)), of which one is the area where the change of internal structure plays a part in the flow, and the other where no such contribution exists.
(ii) The authors advanced an idea stating that the apparent instantaneous change of viscosity accompanied by an additional loading-this paper concerns mainly in such initial stage of the flow-is nothing but the true visco-elastic phenomenon whose domain extends to both sides of t_s curve.
(iii) It was confirmed that the well known time dependency of viscosity discovered by Lillie does not only come from the visco-elastic phenomenon, but also from the change of internal structure with time, which shows that the truce is the phenomenon originating in the superposition of above two causes. The authors have thus succeeded to classify experimentally the phenomena of the viscosity and visco-elastic properties of glass, to which they gave a theoretical interpretation resting on the basis of the “Glassymicro-phase theory” advanced one of the authors (T. Moriya, J. Ceram. Assoc. Japan, 55, 60, 87, 134 (1947); 56, 4, (1948)).
(iv) In order to show a quantitative example a constant representing the time dependency of the change of internal structure was worked out for the glass whose Glassymicrophase had already been unvailed by electron-microscopic observations (M. Watanabe, H. Noake, T. Aiba, J. Am. Ceram. Soc., 42, 593 (1959)).
On the other hand the evaluation of another constant in the term representing the visco-elasticity in Moriya's equation of viscosity (Bull. Tokyo Inst. of Tech., B No. 2, 47, (1955)) was made using the experimental results of additional loading. The theoretical curve thus obtained and the experimental values of the time dependency of viscosity have arrived at a complete agreement.
The same agreement was found in the variety of glasses at several temperatures, so that the authors are of the opinion that the actual existence of Glassymicro-phase and thus the constitution of their theory have amply been justified.

View full abstract

The result of observation on the behavior of oxygen bubbles introduced artificially into four kinds of glass indicated that the bubbles in barium-aluminum-alkali silicate, in sodalime window glass, and in hard borosilicate glass, each fined with oxydizing fining agents, shrank rapidly following the square root of time at the temperature ranging from 1100°C to 1300°C. In the experiment with borosilicate glass fined with sodium chloride, however, the absorption of oxygen was extremely limited and the decrease of diameter was more nearly proportional to the first power of the time than to the square root.
This marked difference may be attributed to the fact that the oxygen potential must be lowered by using the oxydizing fining agents which offer the numerous available vacant sites for the diffusing ions, while such vacancies may not be formed by the physical fining process using sodium chloride. Since the rate of absorption is increased by increases in the temperature of the experiments, it was supposed that a considerable increase in the diffusion constant would surpass the decrease of solubility with rising temperature, and the diffusing oxygen ions may be dissolved in glass by bonding chemically with the elements of variable valencies.

View full abstract

In order to get definit information as to the influence of gypsum on the setting and hardning of portland cement, the authors measured the rate of hydration of three synthetic clinkers and two commercial clinkers using as much amount of water as in the setting and strength test.
The estimation of the rate of reaction was made by the quantitative X-ray analysis to determine the amount of each mineral remaining still unchanged in the hydrated specimens. This method guarantees the results being entirely independent of the nature of reaction as well as the change of form and the composition of the hydrates.
In order to find out the relation between the rate of hydration and of setting of cement clinker, the time of setting of small specimens were measured by the standard method, while the rate of hydration was determined by the X-ray analysis.
The results are summarized as follows:
(1) Among 4 major minerals, tricalcium aluminate phase showed the largest rate giving 10-30% hydration in 3 minutes after the mixing with water. It was found that the addition of gypsum lowered such high values considerably, although the process of mixing up gave more or less the same effect for most of the clinkers.
Without gypsum, however, the clinker of low alite and high tricalcium aluminate phase behaved as flash set, and in this case, the rate of hydration of tricalcium aluminate phase was as high as 30% in 3 minutes after the mixing with water. With the addition of gypsum this value was lowered to 20%, and the clinker became normal set.
The rate of hydration of celite was found to be considerably large, giving 8-20% hydration immediately after the mixing with water, and this value was lowered by several percents by the addition of gypsum. Alite phase was less active giving a few percent hydration after 3 minutes with an increase to 10% in 5 minutes.
Based on above facts the authors concluded that the cause of the reterdation of setting of cement by gypsum would be the decrease of the rate of hydration, essentially, of tricalcium aluminate phase by the action of gypsum and also by Ca(OH)₂ liberated from alite.
The estimation of the degree of hydration of clinker taken all together was made using the rates mensioned above and the mineral composition of clinker calculated from X-ray data with a result of 10% in 3 minutes and 10-15% in 5 hours, in which tricalcium aluminate held the largest value in early stage. Gypsum proved to lower above figures showing a strong trend of checking the hydration velocity of tricalcium aluminate phase.
The authors consider that these facts provide a strong support to the theory advanced by them.
(2) In order to correlate the hydration of cement minerals and the development of the strength of mortars, the test with small specimens (1×1×7cm³, W/C=50%, 1:3 mortar) on one hand, and on the other hand the measurement by X-ray of the rate of hydration of the minerals in the paste containing the same amount of water was carried out.
In a long time alite phase showed the largest rate of hydration reaching to 40-70% in 1 day and to 80-100% in 28 days. Except the very early stage of hydration, tricalcium aluminate phase came in the second place giving the figures of 40-60% in 1 day and 60-80% in 28 days. Celite phase came to the next giving the values, 20-40% in 1 day and 50-70% in 28 days. Belite phase was most inactive, whose rates of hydration were 5-15% in 1 day and about 30% in 28 days.
The addition of gypsum brought the increase of the rate of hydration of alite phase, and in conformity with this, the acceleration of the development of mortar strength throughout the whole period was confirmed. The optimum amount of gypsum was found to be 2.5% reduced to SO₃, except one sample containing exceedingly small amount of alite phase, for which 5.0% SO₃ was the optimum.
The increase of the rate of hydration of belite and celite phase

View full abstract

The results obtained so far by authors have proved that the anodic oxidation was the best for the pickling process of aluminium enamels, since it gave the better effects in adherence, colour and luster, than the chemical pickling or the application of the coating on untreated surface.
Having discussed in a previous paper (T. Yamada, J. Geram. Assoc. Japan, 68, 64 (1960)) the influences of enamels on the different states of metal surface brought out by sevral pickling processes, the author made the direct observations of metal-enamel interface under metal and electron microscopes, the results of which are presented in this reports.
By the anodic oxidation of aluminium a compact and uniform film was formed, which, on the trials changing the time of firing, has proved that the intermediate layer formed by the diffusion of coating into the oxide film is doing much toward the stronger adherence of enamels.
On the contraly, the oxide film formed by a chemical process was rough and uneven containing many cracks. Owing to such structure, the enamel would rapidly worn out, and the enamel would penetrate through the cracks comming into direct contact with metal surface. In such cases it is very liable to get the black coating which easily falls off from the base. Chemical processes make the adherence poor.
The author was able to confirm that anodic oxidatiom was the best for pickling aluminium surface, in the single coating process, since it gave an uniform, compact film which guarantees always “the better adherence.”

View full abstract