Vacuum Web Coating

Free-span deposition has always had the advantage that the deposition can be done across the whole web to the very edge.  This means that the metallized web may not have to be slit to remove any unmetallized edges.  The down side of this process is that there is no substrate cooling and so there can be a problem of the substrate being heated too much. This can mean that the deposition rate has to be limited and the coating thickness limited.  Another advantage of this process is that as there is no frictional constraint to the substrate the web can expand and contract as required and so wrinkling is less likely to be a problem.

The major advantage of using a deposition drum is to extract heat from the substrate and enabling thicker coatings and faster deposition rates to be used.  However there can be a problem of wrinkling because the tension on the substrate holds it hard against the deposition drum and resists the thermal expansion of the substrate which puts the substrate under a transverse compressive force that can result in the formation of wrinkles.

In using a deposition drum the edges of the substrate are masked from depositing material to ensure the deposition drum does not become coated.  This masking does have the effect that the edges of the web are thus not subjected to the heat of condensation or the direct radiant heat from the deposition sources. This means the edges will be cooler than the rest of the web and so marginally shorter and so will progressively take more of the tension as the rest of the web expands as it is subjected to the heat load.  It is not clear if this differential tension is a help or a hindrance to the metallizing process.

This set me thinking to see if there was some solution to this differential heating. One thought was to use the free span idea of full width metallizing but also to use the deposition drum.  In the past when doing research projects I have sometimes had a narrow width of material and rather than have the cost of making a new set of shields I have used a high temperature polymer to wrap around each edge of the drum so that the substrate winds in the gap left between these two polymer shields.  The coating deposits onto the substrate and shields alike and at the end of the deposition run the polymer shields are disposed of and a new set stuck on the drum.  In research this works well as the length of substrate tends to be relatively short.  In a metallizer where there might be 40km or more of the substrate the accumulation of coating on the polymer edge shields could become a problem as the thickness increases towards the end of the roll.  The other concern could be the accuracy of the winding as it is essential that the substrate sits in the gap between the two polymer edge shields. The aim is that there is no gap between the edge shields and substrate so that there is no deposition onto the drum at all.  As rolls can be accurately aligned and rolls can be ground to be concave or convex to provide spreading and centring this should be feasible.

All I need now is to model the two scenarios to see if the full width metallizing using a deposition drum has an advantage over the shielded edge process in reducing the problem of wrinkling.  If this does look to be advantageous I then need to encourage someone to have a go on a real metallizer to see if the reality matches the model.  Until then it is just a thought. If any of you have tried this or have comments or experiences please feel free to write something as a comment.