Background

Thermal spraying is by no means a new technique. It was invented in 1912 in Switzerland and is well established and widespread with coatings produced in this manner being used in all manner of industries. However, so varied are the possibilities afforded by the process that it is continually evolving and growing with new methods and applications. One such evolution is the Purecoat process. A little over three years ago a group of seven interested companies set up a Brite project to find a ways of improving the coating quality of existing techniques, without significantly raising the operating cost. The starting point for such an exercise is to analyze the problems with existing technologies and to do that one has to understand something of the structure of a metal-sprayed coating.
Coating Structure and Buildup via Thermal Spraying

All forms of thermal spraying involve the projection of small heated particles of material at a prepared substrate. Upon impact, the particles spread and freeze, adhering to asperities on the surface as they do so. As more and more particles follow, they land on each other and together form a continuous coating. Upon the properties of each of those particles depends the quality of the final coating. For instance a cold particle will not deform properly at the surface and will lead to porosity around it and a poor cohesive strength. On the other hand if the particle is too hot and exposed to oxygen it will oxidise and the coating will contain a large amount of oxides as well as the desired metal. How can we avoid these problems?
Ensuring Coating Materials are Molten

Arc spraying is a process that uses a wire feedstock melted by an arc. Because of this it is not possible to produce unmelted particles. This solves the first problem of ensuring that all the particles are melted. Now the question remains as to how to stop the oxidation.
Thermal Spraying in Inert Atmospheres

The obvious answer is to spray using an inert gas in an inert gas chamber. We have done this on various occasions since the 1950’s and it works. Even coatings of Titanium can be produced with almost no oxidation. However the use of a chamber means that the components are limited in size and fixed in location and one of the great benefits of thermal spraying may be lost. How much better would it be to be able to shroud the spray in inert gas so as to have a truly portable inert environment for spraying. This idea is also not new. Patents and papers for shrouding of plasma spray systems go back to the 1960’s and varying levels of success have been achieved because it is not as easy as it would at first appear
Thermal Spraying in a Localised Inert Atmospheres

If one just puts the spray head into a plain tube and use inert gas as the propellant then the divergence of the spray stream within the tube tends to coat the inside with molten material, which builds up and finally obstructs the spray. Furthermore, as the fast moving jet of gas exits the tube it creates a region of low pressure and the surrounding air is sucked into the inert gas diluting it to the point of uselessness within a short distance. How can these problems of spray divergence and oxygen entrainment be overcome?

The answers were found by modelling the spraying process by finite element techniques using a Cray Supercomputer.
Modelling

The computer model produced by ESIL in Dublin was able to predict the levels of oxygen in the gas stream at various distances downstream and radial distances from the spray centreline. At the laboratories of CISE in Milan these results were backed up by physical measurements of the oxygen concentrations and a remarkable consistency of results was shown. From these results and knowing from our measurements of divergence that almost all the spray fell within the low oxygen region we were able to predict that the coating material would not be heavily oxidised in flight.

In Figure 1 below, the top illustration is a representation of the concentrations of air and nitrogen in the tube of the shrouding system and immediately outside it. The upper diagram represents a plain tube above showing almost complete dilution with air. The optimised shroud below shows almost pure nitrogen over the particle spray zone.

Prototyping

In parallel with the results from the modelling, prototypes were built to test the divergence and the patterns of flow and even the initial results were very encouraging. We originally set out to beat the plasma specification of 8% oxide content in a Nickel Chromium alloy and actually achieved 3% spraying on tubes. With regard to porosity we aimed for 10% and achieved less than 1%.

What are the practical benefits of this new technology and what can we do that we could not before? To understand this we need to look at the problems involved in some real applications.
Wet Corrosion Resistance

Alloys such as Inconel 625 provide corrosion protection by providing a barrier layer resistant to a whole range of chemical attack. They offer protection by readily forming an adherent oxide film, usually of chromium oxide, on their surface, which inhibits further corrosion. However, this readiness to form oxides mean that normally sprayed coatings will contain oxidised particles, which are therefore depleted in chromium. If the material is depleted in chromium it is clearly not the intended material any longer and cannot be expected to perform as expected. The fact that a conventional coating comprises depleted and non-depleted zones means that it is no longer homogeneous and internal electrolytic cells can be set up causing dissimilar metal corrosion. Research has also shown that these oxide layers around the particles can act as a path for further oxidation and penetration of the oxidising medium to the substrate. Therefore the lower the oxides content of the coating the better. Coating density is also very important in reducing the paths along which the corrosive medium can migrate.
High Temperature Oxidation and Sulphidation.

Components such as boiler firewalls, incinerators and the hot gas paths of gas turbines suffer corrosion at high temperature. The same quality factors arise in the protection of components against oxidation and sulphidation as for wet corrosion but with different materials. In this case high chromium alloys or alloys with chromium and aluminium such as Metallisation 78E and 88E are used. The chromium and aluminium form an adherent oxide scale, which inhibits further corrosion. Where the coating is already oxidised the corrosive species migrate along the boundaries and allow the coating to be corroded internally as well as from the surface.

All of these materials rely on providing a dense resistant barrier of an alloy, which is not degraded by the spraying process. Normal coatings fail due to degradation of the alloy and penetration through the coating. The Purecoat process minimises these deleterious effects leading to better performance and longer life.
Application Testing

Having made our device and seen that the microstructures were promising we wanted to test the coatings in realistic situations. Three application areas were chosen, high temperature sulphidation as is found in boilers burning low-grade coal and orimulsion, wet corrosion such as would be encountered in sea water valves or chemical plant and ductility such as is required in the manufacture of inking rolls.
High Temperature Sulphidation Testing

For the hot sulphidation tests coatings were made on cylindrical samples and subjected to hundreds of hours of testing under a controlled atmosphere in a specially designed furnace. The pieces were examined at 250hour intervals and weighed to discover the weight gain due to sulphide formation. The optimum coatings performed better than an HVOF coating costing 5 times as much to apply.

Thermal Fatigue can be a problem in coated materials and so tubes coated on one side only were thermally cycled for 1000 hours. There was no sign of lifting or spalling of the coating.
Wet Corrosion Testing

To assess the effects upon wet corrosion properties we took one of the most widely used corrosion resistant alloys, Inconel 625, and sprayed it onto carbon steel plates. These plates were tested in two ways. Firstly electro-potential tests were used to compare the sprayed material itself with wrought Inconel 625. Standard Inconel 625 readily passivates at a certain level. The closer to this level that the sprayed coating passivates the more like the wrought material will be its behaviour. Purecoat Inconel 625 behaves very much more like the true alloy than a conventionally sprayed sample.

The second test was a conventional salt spray test. This not only tests the resistance of the material itself but also its permeability. If the salt spray can penetrate the coating then the underlying steel will be preferentially corroded and rust staining will appear on the surface. Testing showed that the Purecoat sample significantly out performed standard spray.