Browsing by Author "Roi, Torsten"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
- ItemFormation and behaviour of bubble curtains in glass melts(Offenbach : Verlag der Deutschen Glastechnischen Gesellschaft, 1995) Roi, Torsten; Seidel, Olaf; Nolle, Günther; Höhne, DiethartIt is widely assumed that bubbles form in melts when their composition and/or gas-dissolving capacity is changed by the dissolution of refractories or sand. This phenomenon is shown by photographs of dissolving sand. In experimental invesdgadons, it was not possible to explain bubble curtains in the vicinity of refractory materials by changes in composition. Computer simulations show that such bubble curtains can be formed even at a relatively low extent of bubble growth in the glass melt. By convective flows, these bubble curtains can be distributed so that they appear to be the result of a boundary layer reaction. At a constant temperature distribution, such bubble curtains maintain their position for a very long time. In the case of heating, the bubbles are distributed over a large volume of the melt. This may suddenly cause the formation of bubbles especially in tank furnaces.
- ItemModeling of the bubble population in glass melts(Offenbach : Verlag der Deutschen Glastechnischen Gesellschaft, 1994) Roi, Torsten; Seidel, Olaf; Nölle, Günther; Höhne, DiethardPrevious studies of refining models have mostly dealt with the mathematical description of the behavior of individual bubbles in glass melts. A further step was the modeling of the ascent and growth of groups of bubbles, including restrictions with respect to the spatial distribution of the individual bubbles. Even though no algorithm was derived in this stage which described the total bubble balance during refining, these investigations have produced important results in the field of bubble growth and ascent. In this study, a general and comprehensive mathematical description of the bubble population during refining shall be given, on the basis of the population balance equation used in chemical engineering. The general balance equation for bubbles during refining is presented together with the corresponding computer model. Several experimental investigations into bubble size distribution are described, together with an analysis of the reduction of the bubble concentration in the pot. The obtained values were used as initial parameters and estimates in computer simulations. Some special results of modeling are shown and discussed.
- ItemΑ characteristic number of refining(Offenbach : Verlag der Deutschen Glastechnischen Gesellschaft, 2000) Nölle, Günther; Roi, TorstenIt was already reported about the prediction of the bubble population behaviour with numerical models based on the population balance model. On the basis of this model, the similarity of analytical function and numerical computation is now presented for some cases of refining. Under these simplified conditions the bubble size distribution of analytical and numerical computation is identical. It is possible to predict the refining time in dependence on the melting depth. In the first step the prediction is possible for laboratory crucibles and pots. The results were compared with experimental values. Some differences between experimental and theoretical refining time were interpreted assuming that the melt volume is homogeneous. This assumption is necessary for these investigations. Α homogeneous melt is found rather in small crucibles than in larger pots. Starting from the population balance model a characteristic (dimensionless) number was found for the refining of glass melts. This characteristic number concept gives some new points of view for a better understanding of the refining process. It shows the temperature dependence of refining time as temperature dependence of glass viscosity and bubble growth rate. The refining capability is described with a simple term. The value of refining capability can be determined by refining experiments. This value can also be calculated if the temperature dependence of glass viscosity and bubble growth rate is known.