Some research into producing a roll to roll coating system for depositing coating by atomic layer deposition (ALD) had resulted in a design whereby the web direction was in the opposite direction to the drum it was wrapped around. The web never touched the drum because the gas supply for the two different process gases and the inert buffer gas all acted as a gas bearing.
When I went back to an early patent on the use of gas injection or a gas wedge to increase the heat transfer coefficient back in 1991they showed not only a rotating deposition drum but also a couple of examples of static drum segments that were cooled and relied on the gas injection to prevent the web being scratched.
There are several problems in using deposition drums. At high speeds the residence time of the web on the drum is short and often is too short to enable the web to reach the same low temperature as the drum. Similarly after the deposition zone there is insufficient time for the web to be brought back to ambient temperature. The deposition drum takes up a large amount of space because of the geometry. The benefit of the gas injection can be variable around the drum as the gas leaks out from the edges of the gap between web and drum. This has been improved by using a porous surface to the deposition drum such that the drum has to be supplied with the cooling liquid and in addition the gas to feed the gap between the web and drum as well as the drum needing to rotate.
If we combine the two different approaches by using the static liquid cooled curved plates but then instead of injecting the gas as the web is about to touch the plate we use the ALD approach and feed the gas through holes arrayed over the whole plate surface or even use a porous plate (sintered powder that remains porous). In this case the gas will float the web and work as an air bearing and the web will never touch the plate surface. This will both save space and also reduce the engineering required as there is no requirement for any drive motor, bearings or any rotating gas or liquid seals. This also opens up the opportunity of changing the positioning of the conductance shields such that the edge shield may also be the conductance shield enabling the leaking gas from the web edges to be pumped by the winding zone pumping and so reducing the contamination into the deposition zone. The space saved could be used to increase the useful cooling available by having strategically positioned cooled plates pre and post deposition zone. Having the cooled plates in the deposition zone the gas throughput could be increased to increase the heat transfer coefficient above that normally obtained in gas injection/wedge systems.
It only goes to show that digging through past patents and keeping up with new developments can lead to some interesting ideas. Actually seeing them used in practice might take a few more years but one can but hope