In a glass making plant, on its way from the furnace, where it is melted, to the machinery where it is dispensed into moulds (molds) or streamed out (extruded) in ribbons for light bulbs or tubes, molten glass passes along a canal, or forehearth, lined with refractory material, probably heated by flames or infra-red lamps to keep it molten and to condition it to uniform and constant temperature for working.

To be able to get better control of the many forehearths in use, the client wanted to build a model of the process with its flowing glass mass flow pattern, and radiative heat transfer from the heaters and by conduction and internal radiation (remember the glass is partially transparent to infra-red) inside the moving stream. While in principle the task is straightforward, the efficient use of computing time to handle all the internal couplings needed some tricks and care. It was also important to ascertain the significance of the interaction between temperature distribution and flow since the viscous flow could be expected to be changed by the strong temperature dependence of viscosity and the Navier-Stokes equations are not well behaved.

Having overcome the misunderstanding at one of the manufacturing sites where they had expected the model to be built physically with flowing treacle (molasses), the work yielded a working model, a masters degree thesis and in the process, eliminated a lot of industrial folklore.

For another example in the glass industry look at glass moulding.

If you have a problem in the dynamics or control of a distributed parameter system like this forehearth... moving streams with heat transfer or other changes going on (like chemical reactions for example).... then contact me.