Parent Category: Industries

Simulation of Glass Floating

The simulation of glass floating can be performed with our CFD software NOGRID points. Float glass is produced by floating a continuous stream of molten glass onto a bath of molten tin. The molten glass spreads onto the surface of the metal and produces a high quality, consistently level sheet of glass that is later heat polished. The glass has no wave or distortion and is now the standard method for glass production and over 90% of the world production of flat glass is float glass. The float glass process was developed by Sir Alastair Pilkington and patented by Pilkington in 1959 and the detailed history of the development is process is described by Sir Alastair Pilkington in his review lecture to the Royal Society of London in 1969 (Pilkington, L.A.B. Proc. Roy. Soc. London 1969, A314, 1-25).

Computing float process with meshless CFD

Figure 1: Float process

If molten glass is poured onto a bath of clean molten tin, the glass will spread out in the same way that oil will spread out if poured onto a bath of water. In this situation, gravity and surface tension will result in the top and bottom surfaces of the glass becoming approximately flat and parallel. The molten glass does not spread out indefinitely over the surface of the molten tin. Despite the influence of gravity, it is restrained by surface tension effects between the glass and the tin. The resulting equilibrium between the gravity and the surface tensions defines the equilibrium thickness of the molten glass (T).

The glass comes from the furnace and flows over a 'dam' or spout-lip where the continuous stream of molten glass flows onto the bath of molten tin. The stream of glass is pulled along the top of the molten tin by haul-off conveyors at the end of the float area which transport the glass into the annealing area. Figure 2 shows the results of the complex free surface flow of the glass, which is poured onto the tin bath. The tin flow itself is not taken into account here, but we considered the buoyant force between glass and tin. The result is that the glass is able to dip under the tin level. Especially in the vicinity of the lip it dips deeper in the tin bath.

Free surface flow in the spout-lip area

Figure 2: Results of the complex free surface flow in the spout-lip-area

At the start of the float area the molten glass spreads outwards with flat top and bottom surfaces and the thickness decreases towards the equilibrium thickness (T). The thickness can then be further controlled by the stretching effect of the conveyors as it cools until it reaches a certain temperature when it exits the float area and enters the annealing lehr. Whilst the equilibrium thickness is approximately 7 mm the process has been developed to allow the thickness being controlled between 0.4 mm and 25 mm. For thin sheets, the exit conveyor speed can be increased to draw the glass down to thinner thicknesses. This drawing down will also result in a decrease in the sheet width and to prevent unacceptable sheet width decreases edge rolls are used. Edge rolls grip the outer top edge of the glass and do not only reduce decrease in width but also help to reduce the thickness even further.

For thick sheets, the spread of the molten glass is limited by using non-wetted longitudinal guides. The glass temperature allows the spread remaining uniform and is reduced until the ribbon can leave the guides without changing dimensions.