THE REVERSIBLE THERMAL EXPANSION OF REFRACTORY MATERIALS

Abstract

The reversible thermal expansion from 15–1000°C was measured for kaolin, siliceous and aluminous fire clays, quartzite, alumina, magnesia, and carborundum, after preliminary burnings at cones 06, 9, 14 and 20, and as well as for English commercial silica bricks before and after use in a coke oven and the roof of a steel furnace. Kaolin and bauxitic fire clay after calcination have a regular reversible thermal expansion which does not vary much with the temperature of calcination. Siliceous fire clays, after calcination at cone 06 (980°C) or cone 9 (1280°C) display irregularities (departures from uniformity) in their expansion. Between 500° and 600°C they show a large expansion due to contained quartz and on cooling the contraction in that region is larger than the corresponding expansion. Moreover, the expansion between 100° and 250°C after being fired to cone 9 (1280°C) exceeds the average. After calcination at higher temperatures, cone 14 (1410°C) or cone 20 (1530°C). these materials gradually lose these peculiarities until on incipient vitrification a linear expansion similar to that of kaolin is attained. This change is due to the destruction of quartz by its interaction with the clay material and fluxes; it takes place most easily in a fine-grained, rather friable clay such as ball clay. The previous thermal treatment necessary for a particular clay in order to obtain regular expansion in use can only be determined by trial. It can be stated with confidence that in such a piece of apparatus as a glass pot or crucible, a distinct gain will result from maintenance at a high temperature for some time before use, but that the red heat of an ordinary pot arch is useless for the purpose. An increase in the porosity of a fire clay was accompanied by a corresponding decrease in expansion between 15° and 1000°C until a porosity of 50% was attained. Further increase in porosity produced very little change in the expansion. No irregularities in expansion were shown by magnesia brick, carborundum, or alumina bonded with 10% of ball clay. Welsh quartzite with lime bond, either unfired or after burning at cone 06, had a large expansion to 550 °C and a much larger expansion from 550–600 °C due to the inversion of α to β quartz while from 600–1000°C a slight contraction took place. Firing to cone 9 converted part of the quartz into cristobalite, thus increasing the expansion from 200–250°C. This conversion was considerably increased on burning for two hours at cone 14, which greatly reduced the expansion from 550–600°C with a corresponding increase of that from 200–250°C. The conversion of the quartz into cristobalite was completed by a further heating for two hours at cone 20. Determinations of refractive indices and specific gravities confirmed these results. Flint inverted to cristobalite with greater ease than quartz. Commercial silica brick consisted chiefly of cristobalite and unconverted quartz and showed a large expansion up to 300°C, followed by a considerably smaller but regular expansion to 550°C. From 550° to 600°C the rate of expansion was considerably increased, but above 600°C the change in dimensions was small. The innermost exposed layer of a silica brick after use in a coke oven was an impure glass with a steady expansion, but only half as large as that of the layers of brick behind, which was made for shelling away. A silica brick after use in a steel furnace was divided into four layers. The layer exposed to the furnace heat was practically all cristobalite and silicates, the next layer the same, the third layer showed some α to β quartz expansion as well as the α to β cristobalite expansion, while the fourth (outermost) layer exposed to air was similar to the brick before use. In these bricks exposure to high temperature had evidently completed the change from quartz to cristobalite which had been largely effected in the kiln during manufacture. Little or no tridymite had formed. The reversible thermal expansion from 15–1000°C of the commercial silica brick examined was 1.1 to 1.3%, about double that of fire clay brick.