The Viscosity and Working Properties of Glass. Part II. The Rate of Cooling and Setting of Colourless and Coloured Glasses.

JSGT 1943 V27 T094-T112

As a preliminary to studying the more complicated problem of ascertaining the changes in temperature and viscosity occurring at different points in a mass of glass undergoing conversion into glassware by hand or machine manipulation, a comprehensive study of the rate of cooling of glasses contained within a crucible has been carried out by methods originated in this Department and followed by subsequent investigators whose work is referred to. The glasses tested comprised five commercial types, colourless and coloured, based on a soda–lime–silica composition, two colourless alkali-lead oxide glasses, a high softening point boroaluminosilicate and a soda–silica glass. Their compositions and viscosities are fully set out in Part I of this investigation (this J., TRANS., 1942, p. 215). In addition, eight glasses, coloured blue by cobalt oxide, present in varying concentrations up to 5%, were studied. Determinations were made by platinum-platinum rhodium thermocouples of the temperatures at the glass surface and at intermediate points within the interior when the crucible and the glass within it were cooling under two different sets of conditions. From the temperatures recorded at different times of cooling the corresponding viscosities were deduced from the data given in Part I of this investigation. The temperature differences between the surface and interior of a mass of glass during cooling depend on the time during which the cooling has proceeded; on the initial temperature from which cooling begins; and on the composition of the glass, particularly the presence of colouring constituents. As regards the effect of time, the difference in temperature between surface and interior increases rapidly to a maximum during the early stages of cooling and then slowly decreases to zero as the glass cools uniformly to room temperature. In regard to the effect of the initial temperature from which cooling begins, the relative times required for the soda–lime–silica glass No. 9 to cool through the first 400° from the different initial temperatures were : (a) for the surface, 1400–1000°, 1·0; 1300–900°, 1·43; 1200–800°, 1·85; (b) for a point 0·8 in below the surface, 1400–1000°, 1·0; 1300–900°, 1·29; 1200-800°, 1·60. These data apply to conditions described, in which the crucible and glass cooled slowly in an insulated chamber. Data are given in the paper for other glasses and other temperature ranges and conditions. The differences in temperature between surface and interior were markedly greater in the coloured than the colourless glasses, partly due to the greater emissivity of the former and partly due to the selective absorption they exercise, resulting in the longer retention of heat in the interior during the cooling process. When cooling from an initial temperature of 1400°, the temperat1-ues, after an interval of 2 minutes were, for a colourless soda–lime–silica glass (No. 9), a carbon-sulphur amber glass (No. 16), a cobalt blue glass (No. 22) and an iron-manganese green glass (No. 20) : (a) at the surface, 896°, 804 °, 820° and 7 53°, respectively; (b) at 0·8 in. below the surface, 996°, 1036°, 1045° and 1063°, respectively. In coloured glasses the temperature differences between surface and interior are very marked within a narrow layer at the surface. Thus, whereas between the surface and 0·1 in below it the difference was only 24° in the case of colourless glass No. 9, the differences for the same small depth for the coloured glasses Nos. 16, 22 and 20, were 79°, 59° and 102°, respectively. The rate of setting of a glass is influenced by the temperature coefficient of viscosity and by the factors previously mentioned, namely, the initial temperature from which cooling begins and the composition, particularly the presence of colouring constituents. Tables and graphs show in detail the effect of all three factors. The very marked differences of temperature within the first 0·1 in of the surface layer in the case of coloured glasses gives rise to correspondingly big viscosity differences resulting in a "skin" effect which governs subsequent manipulative operations. The effect of other factors is also referred to.

James Boow & W. E. S. Turner