Читать книгу Encyclopedia of Glass Science, Technology, History, and Culture - Группа авторов - Страница 100

In spite of the simplicity of its principles, the float process is not readily implemented because flow in both tin and glass and heat transfer among tin, glass, and the radiative field are really complex processes. Glass forming is mainly determined by parameters such as the glass flow rate, conveyor speed, rotating speed and angle of top rolls, and by the viscosity distribution within the glass ribbon. It goes without saying that the viscosity of the glass strongly depends on temperature, but the temperature distribution in the bath is itself influenced by radiative heat transfer, the flow of molten tin, and the glass forming conditions. Radiative heat transfer is predominant in the bath at temperatures higher than 600 °C, but the flowing molten tin also contributes markedly to heat transfer as a result of its high heat capacity, high thermal conductivity (about 50 times higher than that of the glass), and low kinematic viscosity (about 8 times lower than that of water), whereas the glass flow carries a large amount of convective heat. On the other hand, molten tin flows by traction from the glass ribbon (i.e. velocity distribution) and buoyancy convection (i.e. temperature distribution in the bath). In order to understand forming conditions, one thus needs to understand the whole set of processes taking place in the bath since glass forming, heat transfer in the bath, and flow of molten tin affect one another in a very complex manner.

As summarized by C.K. Edge [4], the float bath thus is “a remarkable entity which, although first envisioned as a finisher of glass surfaces, also functions as a container, a conveyor, a forming unit, a chemical reactor and a heat exchanger.” In view of this complexity, valuable information has been drawn from mathematical simulations not only of the glass forming mechanisms, but also of the temperature field and the mutually related dynamics of the molten tin and glass ribbon. As examples, calculations with finite‐element methods of the thickness contour over the glass ribbon and of the lateral thickness distribution of the ribbon at the bath exit are shown in Figures 11 and 12 [10]. The rather good agreement of such model values with the temperatures, thicknesses, or ribbon shapes that can be measured on line illustrates how simulations can be used to optimize the operating conditions, to design new facilities, and to check new ideas for process improvement and development.

From a chemical standpoint, potentially annoying impurities are oxygen from leaks or the N₂–H₂ gas mix, and sulfur originating from the molten glass. Both deteriorate productivity and glass quality if they induce alteration on the bottom surface of the glass caused by reactions with tin, and contamination adhesion on top and bottom surfaces (Chapter 5.6). In the so‐called oxygen and sulfur cycles (Figure 13), SnO and SnS vapors form, condense, and precipitate, the former as SnO₂ and the latter as SnS (which is ultimately reduced to Sn particles). Thanks to extensive research, however, these impurities are now carefully managed through monitoring, sealing, cleaning, atmosphere controlling on flow and pressure, etc.