Cooking up a Coating: Control of Key Process Variables

I remember sitting down once with a mental challenge of writing down the variables in the thermal spray process that could have a significant effect on the deposit being produced. After about two or three minutes, I had reached a total of 47, and I thought that was enough for a first effort.

What this means, of course, is that there are many factors to consider and control when attempting to produce a quality coating. Over the 40-plus years I’ve been involved in this process, I have witnessed significant strides in improving robustness, including advancements in mass flow controllers, PC-integrated systems, and diagnostic sensors. This is all fantastic stuff, but as thermal spray is not yet in the realms of CNC machining, the level of manual intervention still required means that some of the key techniques used back when I started in this business are still very relevant today.

In this article, I hope to provide an insight into the “recipe” of variables affecting air plasma spray (APS) and high-velocity oxygen fuel (HVOF). While I appreciate that these are only two of the many thermal spray coating processes out there, for brevity’s sake, it’s probably best to concentrate on a couple of the favorites.

The Coating Recipe

Just like baking a cake, any coating needs a starting recipe to make sure there is a good chance of getting what you expect when the coated component comes out of the “oven.” For thermal spray, the usual format follows that of a spray parameter sheet — Fig. 1. Here, the idea is to control as many of the input parameters as possible to keep variability to a minimum. In this case, the parameter sheet calls up the material to be sprayed, the gun hardware to be used, and the system input parameters required. Figure 1 also provides an indication of the resultant microstructure. Alongside this structural expectation, there will most likely be a set of coating properties that will need to be met. These will often be controlled by internal or customer test specifications.

The Starting Ingredients

Both APS and HVOF use powder as a consumable. Depending on the powder chemical composition and the required coating properties, powders can be manufactured using a wide variety of methods — Fig. 2. The manufacturing method used will have an impact on powder density, morphology, phase distribution, etc., and therefore also on the resultant deposit.

Once made, powders are subsequently sized to suit the process being used. Figure 3 shows typical sizing for a range of powder-consuming systems. For both APS and HVOF, the powder can be subsequently optimized to produce coatings with specific properties. For example, if an as-sprayed coating is required with a low surface roughness, then it makes sense to select a powder sized toward the bottom end of the usual size range. Of course, there are always other factors to be aware of. In this case, choosing a material that is too fine can cause hardware blocking issues for both APS and HVOF processes.

Overriding Principles

Once we have chosen our powder and thermal spray process, we need to carefully consider what happens when the two interact. The goal is to produce a functional coating, but how do we ensure we get what we want?

I already mentioned the 47 variables, but although this may seem like a long list, many of these have an influence (admittedly sometimes on multiple levels) on just two key factors: particle temperature (T) and particle velocity (V).

I always try to imagine myself as a particle in the spray plume and what external forces are acting upon me (strange, but true). The choice of hardware used (size of nozzles and powder injectors), flow of gases, current settings, etc., all influence powder particle thermal and kinetic energy levels, and therefore, the coating produced.

This overriding principle of particle temperature and velocity is not only the cornerstone of spray parameter control but also forms one of the fundamentals of flame sensing technology. As indicated in Fig. 4, infrared emissions from particles in the flame can be utilized to provide data on these two principal output variables. The general theory is that, if we can reduce variability in defined values of particle T and V, then we should end up with the coating we are expecting.

Picking the Correct Utensils

The working end of the thermal spray process is usually referred to as the spray gun or torch. Before we even think about what gases and power levels we are going to employ, we need to set the gun up with the right hardware.

When preparing an APS gun, hardware can be process gas specific. The choice of primary and secondary gas will affect plasma energy levels, and items such as nozzles (anodes), electrodes (cathodes), as well as parts such as gas swirl rings, will have to be chosen to reflect the gas combination used.

Returning to our T and V philosophy, the size of the nozzle bore (which can be the controlling orifice in the system) will have a profound effect on gas and, therefore, particle velocity. It will also affect the time the powder particles spend in the plasma “flame” (particle dwell time) and the amount of heat transferred.

The control of delivery of powder into the APS plume is also significant. The powder feed rate and injection methods are key to the process. As can be seen in Fig. 5, the choice of powder port/injector size (defined by through-hole diameter) as well as powder carrier gas flow will influence the position of injection in the plume. Nonoptimized powder injection usually leads to a nonoptimized coating (unmelted particles, nonuniform oxides, etc.). In addition to physical changes to carrier gas flow rates, leaks in the system, injector blockages, and wear will also affect injection. Powder feed system leaks are one of the more significant root causes of reported coating issues.

The positioning of the powder port in terms of distance from the plume as well as injection angle will also have an influence on particle heat transfer. In particular, powder port angles can be chosen to help control particle dwell time. It makes sense that a material requiring more heat input to soften would prefer to have a longer dwell time. On that basis, higher melting point materials such as ceramics tend to be injected backwards into the plume (for example, using a #3 port as shown in Fig. 6). The reverse is of course true for lower melting point materials.

The hardware choice for HVOF systems tends to be a little more straightforward but still follows the particle T and V trend. Once the design of the gun and choice of fuel have been established, the major hardware choice tends to be based on barrel length as well as particle dwell time and acceleration — Fig. 7.

Design of barrels (HVOF-LF) or extended aircaps (HVOF-GF) can also influence particle dynamics. Many HVOF gun designs utilize a de Laval (convergent/divergent) nozzle, which accelerates the hot, pressurized gas passing through it to a higher supersonic speed by converting the heat energy of the flow into kinetic energy.

Stirring the Mix

Irrespective of the process chosen, consideration of what happens when the powder enters the process is significant.

Powder feed rate is a key variable. Whether it is HVOF or APS, the process flame will have a fixed enthalpy level for a fixed parameter set. If the amount of powder delivered to the T and V source varies, then so will the heat transfer. This can have a profound effect on all coating properties. Proper management of powder delivery within control parameters is significant.

In addition to routine process control, the design of modern powder feeders helps reduce variability in powder feed rate — Fig. 8. As well as accurate material delivery systems, closed-loop “weigh as you spray” feeders help ensure that the desired feed rate is kept tightly within the defined tolerance.

Once the powder has entered the flame, again dwell time becomes a controlling factor. The time taken for the powder particles to strike the surface is affected by gas flows and hardware choice, as well as significantly by spray distance. As can be seen in Fig. 9, a complex interaction of powder size distributions will affect T and V conditions. The optimum particle properties do not, therefore, always coincide with the determined spray distance. Quite often, it is a compromise that can be overruled by part geometry and access issues.

Setting the Dials

Both APS and HVOF processes are enormously active thermodynamic systems. If we are generating 40 kW or more of energy at any one time, there must be a good reason for it.

A quick calculation shows that an HVOF powder particle traveling at 750 m/s over a spray distance of 375 mm (14.76 in.) takes less than a thousandth of a second to hit the substrate after leaving the end of the gun barrel. Not surprising then, that a great deal of energy (thermal and kinetic) is required to create the required coating.

HVOF processes rely on the combustion of fuel with oxygen to generate flame temperature and velocity. The amount of gas (and/or liquid) fed into the gun needs to be regulated to ensure optimum combustion occurs. This regulation within controlled tolerances is usually carried out via flowmeters (the latest being digital mass flow). Parameters chosen for any given material are not necessarily based on maximum gas temperatures or complete (stoichiometric) combustion. In many cases, they have been empirically defined to produce the desired coating properties. That said, once developed, some clear rules exist that can be used to modify spraying conditions. Figure 10 shows a typical flowchart for an HVOF-LF system. Here, we can see how changes in flow rate and ratio affect flame (and therefore particle) temperature and velocity.

The diagram also indicates the availability of measured HVOF gas combustion pressure. This process feedback value can be enormously valuable in providing data on the health of the system. Suitable tolerances applied to the pressure value can indicate problems with hardware blockage or incorrect combustion conditions — a useful warning of problems on the horizon.

Moving on to APS, plasma spray is an electrical process, and its energy is derived from the ionization of gases. The amount of gas used helps define particle velocity, while the types of gases used determine energy levels.

Figure 11 is a historically well-viewed chart, but it provides a good indication of energy levels available in typical gases used in the APS process. The selection of primary and secondary gases will certainly affect the transfer of heat to the powder particles and consequently influence the coating properties and deposit efficiency. It makes sense, for example, that you would likely choose N2 /H2 (primary and secondary) gases when depositing a high melting point material such as yttria stabilized zirconia (e.g., Metco 204 NS) and Ar/He when spraying a temperature sensitive material such as Co/WC (e.g., Metco 73F NS-2).

Referring to our parameter sheet, detailed in Fig. 1, you can see that process gas flows and plasma energy levels need to be controlled. The latter is primarily defined by electrical power and as mentioned previously, is reported in watts (typically kW) calculated as current (A) multiplied by voltage (V). The voltage is produced as a result of the ionization process and is a good feedback tool for monitoring system robustness. This is especially true for traditional single electrode APS systems as the voltage (and therefore the power) tends to degrade as consumable hardware ages. This is not so much the case with cascaded plasma systems.

To help control the APS process, the defined parameters are entered into a process controller, such as that offered by the MultiCoat™ 5 user interface — Fig. 12. This interface links with a range of (in many cases) closed-loop devices that control and maintain set parameters within defined tolerances.

This type of interface is our primary tool for transferring the parameters that may have been supplied to us by the customer. Input of amperage, primary and secondary gas flows, carrier gas flows, and powder feed rates all transferred from the parameter sheet will provide control of a significant number of key process variables. The latest controllers offer features for setting and monitoring tolerances. Alarm warnings can be a really useful feature to ensure that any process drift is actioned before potentially affecting the coating (and incurring unexpected costs).

Despite technology, a good deal of attention is still required to choose the right powder, properly set up the gun hardware, set the correct spray distance, etc. (as well as considering the remainder of the 47 variables I’ve not had time to mention) to make sure we get the desired coating and all its required properties.

Tasting the Cake

The goal of all this effort is to produce our coating “cake” and make sure that the customer enjoys the “taste.”

Conclusion

The aim of this article has been to provide an overview of the parameters that need keeping a careful eye on while baking our coating cake. A keen focus needs to be applied to all the variables that can affect the thermal spray process and great care needs to be taken to use a variety of methods to make sure they do not drift outside defined tolerance bands.

The use of parameter sheets is a great method to make sure our coating recipe is well defined, but setting up and keeping a proper eye on the process is invaluable in ensuring the expected coating with the desired properties is produced.

STEPHEN BOMFORD (steve.bomford@oerlikon.com) is Customer Solutions Centre manager, Oerlikon Surface Solutions, United Kingdom.