In the thermal plasma spraying process, the coating material, in fine powder form, is injected into a high-temperature plasma jet. The powder particles are melted, and the droplets formed are projected at high speed towards the material to be coated (substrate).
To get a good even coating onto the surface of the substrate, each droplet on collision has to slow down, spread evenly into a pancake shape (called a splat) and strongly adhere to the substrate surface. Once an initial layer of splats has been laid down, further droplets must splat such that they successfully cohere to this layer and within the coating.
With some surface/droplet combinations, instead of the droplet forming a disc-shaped splat on collision with the substrate, the result is a splashed splat. If a coating contains splashed splats, its properties, such as electrical resistance, thermal insulation and wear resistance, could well be compromised. In addition, the coating may peel off under heavy load. Any type of coating failure inevitably results in customer dissatisfaction, which could mean an end to that line of business for an industrial spray company.
Some of the models set up to explain the mechanical processes occurring during splat formation were, in Margaret’s case, found to be wanting. What her research team was seeing was splats when the model predicted splashing and vice versa.
Given that Margaret’s background was in surface chemistry, she immediately thought that the surface of the substrate must be playing a role in determining whether the droplet splats or splashes. In the microseconds that it takes for the droplet to hit the substrate, spread and then solidify, Margaret has found that there is a surprising amount that happens both chemically and physically.
In the case of coating an aluminium substrate with nickel/chromium alloy, it was discovered that the chemistry of the surface of the aluminium played a key role in causing a splashed splat. With the aluminium substrate held at room temperature, splashed splats were seen, whereas with the substrate held at 350°C, good splat formation was seen.
Margaret’s investigation of this established that, at room temperature, the native oxide layer present on the aluminium surface consisted of several thinner layers of various chemistries. The inner layer was pure aluminium oxide, but the outer layers were hydrated oxides and hydroxides of aluminium.
When molten droplets at a temperature of about 2500°C hit this surface, the rapid heating of the hydrated aluminium oxide/hydroxide layer leads to its conversion to ‘dry’ aluminium oxide and water vapour. This escaping water vapour disrupts the spreading of the droplets into splats, causing them to splash.
When the aluminium substrate is heated to 350°C prior to coating, the hydrated aluminium oxide/hydroxide layer is decomposed into ‘dry’ aluminium oxide due to the loss of water. On colliding with this now ‘dry’ surface, the droplets form splats with little or no splashing.
A key finding that this research revealed was that, in plasma spraying of coatings, the surface chemistry of the substrate could have a major influence in the droplets splatting or splashing. Heating the substrate prior to coating may be one way of favourably altering the surface chemistry, thus enabling a more consistent droplet-to-splat conversion rate. This in turn will enhance the adhesion of the coating to the substrate as well as the cohesion of the splats within the structure of the coating.