In order to investigate some preventive measures against the oxidation of silicon carbide refractories the sintered specimens of fine SiC powders bonded by Kibushi clay, a popular binder in Japan, were heated at 1175°-1450°C for 50-100 hours in the current of dry oxygen to measure the weight increase, apparent porosity, length and the expansion up to 500°C. The results were compared with those obtained with the bondless bodies formed by adding a small amount of Syton 2X, a silica gel preparation. They weres ubjected to oxydation under the same conditions (as before).
Based on the following experimental results, and on some additional considerations the author advanced a tentative plan of improving the characteristics of clay bonded refractories.
(1) The bondless body used in this experiment was liable to oxidation with a large volume change. But, at the temperatures higher than 1400°C a small amount of impurity promoted remarkably the lowering of apparent porosity during the process of oxydation, so that the oxydation itself was held back to some degree.
(2) Sintered body with 15-30% Kibushi clay as binder showed only a small volume change after long hour's heating at 1300°-1350°C or higher, because of the slow rate of oxidation. At a lower temperature (1175°C) the crystallization of the oxidation product, SiO₂, and the influence of cristobalite liberated from the clay were corelative to show a larger expansion than bondless bodies. At higher temperatures (above 1400°C) a melt of low viscosity came out from the body making the surface more porous. The author concluded that the increase of clay substance gave practically no merit of preventing the oxidation of SiC. Taking into other important points the amount of clay should be as low as possible.
(3) Kibushei clay used as bond failed to prevent the oxidation at the temperatures lower than about 1230°C. The bodies of low porosity produced by maturing sufficiently at a temperature higher than 1450°C the material added by the components such as Na₂O and K₂O, which effectively glassify SiO₂, may be expected to improve the defects at least to some degree.

View full abstract

In the literature concerning the deterioration of steatite (H. Thurnaure, A. R. Rodriguez, J. Am. Ceram. Soc., 25, 443 (1942); W. R. Foster, J. Am. Ceram. Soc., 34, 255 (1951)) the growth of magnesium metasilicate formed in the body, or the transformation of protoenstatite comming out in the cooling period were said as the principal sources of this defect. However, there is no example of discussing in detail on the influence of glass phase coexisting in the body.
The author has carried out the investigations on the characteristic properties of MgSiO₃ with and without glass phase in high temperatures using X-ray diffraction, differential thermal analysis, dilatmetric method and electron microscope. The MgSiO₃ without glass phase showed the same characteristicsof transformation as pointed out by W. R. Foster (l. c.), while the MgSiO₃ with glass phase (protoenstatite), when quenched in some degree, was readily super cooled to secure the stability at an ordinary temperature.
Based on his experimental results the author discussed the deterioration of steatite pointing out that the devitrification in glass or liquid phase caused by the shift of phase equilibrium in the body in the cooling period plays the leading roll in this phenomenon.

View full abstract

A series of experiments was carried out with an idea that better blast furnace cement may be produced by the combination of the portland cement of normal fineness with finer blast furnace slag. The granulated slag was ground to nearly 5000cm²/g of Blaine fineness to mix, in different ratios, with the portland cement of 3000cm²/g. Using the mixture and normal sand the specimens of 1:3 mortar of the size 40×92φmm were formed, and after water curing the water absorption under the pressure of about 3mm of H₂O was measured with an apparatus of the author's own design. Similar experiments were carried out, using the slag of the fineness of about 3000, 4000 and 5000cm²/g in order to know the effect of grain size. The strength of the mortar specimens 40×40×160mm in size were also tested.
The following results were obtained:
(1) Blast furnace cement containing 60, 70 and 80% slag of the size of about 5000cm²/g showed the absorption as low as 3-15% of pure portland cement after the age of 2 or 3 weeks.
(2) The effect of grain size on the absorption fell gradually with age.
(3) The absorption decreased with the increasing fineness of slag.
(4) In early ages the difference of absorption between two cements containing 3000 and 4000cm²/g slags was far larger than those between 4000 and 5000cm²/g slags.
(5) The difference of absorption caused by the fineness of slag decreased with the increasing age of specimen.
(6) Compared with portland cement blast furnace cement showed a smaller ratio of absorption/evaporation telling that it had a higher drying rate than pure portland cement notwithstanding the lower surface of the specimen was always in contact with water.
(7) Furthermore, it was noticed that the blast furnace cement gave a larger amount of bleeding water and a smaller content of air when filled in mould.
(8) The blast furnace cement of 5000cm²/g powder gave higher two weeks strength than portland cement.

View full abstract

The differential thermal analysis of montmorillonite and other clay minerals at low temperatures, extending from room temperature to -195°C, was carried out using dry ice and liquid nitrogen as refrigerants.
Montmorillonite gave three peaks of different type. The first peak at around -4°C would be due to the freezing of mechanically mixed water, while the second one at about -7°C is probably due to the freezing of the adsorbed layer composed of liquid water. The third peak comming out between -20° and -30°C may be attributed to a part of liquid water being in existence in the interfacial layer between the layer lattice of montmorillonite.
The first and second peak were affected by the degree of hydration, the size of particles and the condition of thermal analysis. The third peak was small and low compared with other ones. The temperature and the hight of the peak changed with the cations held between the layer lattice.
From the equilibrium diagram of water the author was able to estimate the pressure, under which the water film between the lattices may exist as liquid should be as high as 2000atm.
The differential thermal anlysis of attapulgite, nontronite, vermiculite and also hydrated halloysite gave three peaks somewhat similar to montmorillonite. However, in kaolinite, illite, and Al₂O₃ only one peak was observed at about -2°--5°C which probabey is due to the mechanically mixed water.

View full abstract