Metal is deposited on a moving web of polymeric material under vacuum.
In the process of converting polymer sheet to metalised material for packaging, the usual method is do it by a vacuum web coating process. The method is also used for other applications. I have, over the last few years, developed a series of models of vacuum web coating processes.
Repository of Knowledge.
One model provided the user with a working simulation of the plant, developed while the plant was being built and commissioned. It was based on analysis of the underlying physical phenomena and became a repository of knowledge about the plant. Being a dynamic and interactive model (simulator) it could be used for training operators, trying out operational procedures, helping with diagnosis of irregularities and assesing the effects of design or material changes. It was set up to run on any personal computer and allow the user to drive the plant in much the same way as operating the real thing, but quicker and without the risks and expense.

Tension around a Drum
Dr Dilwyn Jones and I have been working on getting a good model of the development of the tension distribution around a coating drum, subject to cooling, heating and friction. Because the tension at the entry and exit can be controlled by the rest of the winding system, the distribution around the drum is a two point boundary value problem. We have some workable approaches. Let us know if you are interested.

Wrinkles?
Another model looked at the development of tension round the cooling drum as the polymer film was heated and, having been tensilized (pre-stretched), tried to shrink back towards its original size. The range of laydown and takeoff conditions was explored to see how the controllable conditions might affect the formation of wrinkles while on the drum.
In a later investigation, we (McCann, Jones, Affinito) looked at the mechanism of buckling in cross machine direction due to thermally induced compressive strain and created a new model for the mechanism of buckling of a thin cylindrical shell off a drum.
These models have been reported at Society of Vacuum Coaters (SVC) technical meetings. (see list of publications)

Evaporation, Thermal Models
In the course of this work, models of the physics of the evaporation of the coating material (metal) and its flow as a gas to the web on a cooling drum have been created. Such models included everything from power input, radiation convection and conductive heat flux and choked sonic gas flow to coating thickness. Tools used included interactive simulations programmed in Borland's Pascal (Delphi) and later others using Mathcad. I have also begun to develop a tool in SciCosLab (open source, free) to allow anyone to assemble a thermal model of a whole coating process from unwind to windup, simply by concatenating function calls.

Evaporation from Liquid Surface.
In collaboration with Dr John Affinito and Chris Sheehan of Moltech Corporation (now Sion Power), a model of the evaporation process from a liquid surface to the interior of a chamber from which it escaped has been devised to explain the internal heat fluxes and material transfer that actually occurred. (See list of publications, 14th Bakish conference, 2000.) an extension of this model to show the dynamic heat transfer and web temperature, allowing for the changing view of the evaporative source as seen from the web was presented at the 2001 Technical Conference of the Society of Vacuum Coaters (Philadelphia, PA, USA, April 2001).

Data Mining.
In collaboration with Dr C.A.Bishop, Bishop Consultancy Ltd, UK, an investigation of the application of Chemometrics to Web coating issues, using "Pirouette" software was undertaken and reported in the SVC 42nd Technical Conference, Chicago, April 1999 (see list of publications). We found that the software did a good job of identifying key sources of variations without the need for real statistical skills.