An Experimental Investigation of Alumino-Silicate Refractories of High Purity, for Use in Glass-Melting

JSGT_1950_V34_T274_T304

A study has been made of the corrosion resistance of refractories of high purity when exposed to molten glass of ordinary colourless bottle glass composition. The refractory mixtures were made mostly from kaolin, alumina and silica in various proportions; but in one or two cases burnt kaolin (Molochite) was used in place of some of the raw kaolin, and in one series sillimanite was used as the “grog.” The compositions were varied to give alumina/silica ratios ranging from 30:70 to 80:20. For most of the experimental refractories the grog was prepared by firing a suitable mixture of the raw materials, the grog being subsequently bonded by a similar mixture. The proportions of grog and bonding mixture were varied, the proportion of kaolin in some of the mixtures being 40%, and in the others 35%. The shrinkages during firing and the apparent porosities of all the fired refractories were determined; the effects on porosity produced by firing to different temperatures and by holding the mixtures at particular temperatures for longer or shorter times were also studied. The corrosion resistances were compared by determining the extent to which cylindrical test-specimens were corroded after partial immersion in molten glass at 1400° and 1500° for periods of 10 h, the specimens being moved slowly through the glass. To determine the effect of grog size on corrosion resistance, tests were made using mixtures containing different grog-gradings but otherwise identical. The results indicate that corrosion resistance cannot be correlated in any direct and generally applicable way with overall chemical composition, or apparent porosity, or, again, permeability to molten glass. With certain provisos, however, ·the conclusions can be drawn that corrosion resistance is enhanced by the use of fine-grained grog and of lower proportions of raw clay to grog, also by firing the grog and the refractory mixture at a sufficiently high temperature for a sufficient time. The difficulty of formulating any useful specification other than a “performance specification” is pointed out, and it is suggested that the corrosion test used in the investigation should be given full consideration in the event of such a specification being put forward.

H. Moore & R. Heeley