Commercial thermal barrier coatings consist of three layers. The first layer is typically an aluminum-rich bond coat that is based on the compound nickel-aluminum, or NiAl. The bond coat is applied directly to the turbine blade. The second layer is a thin, thermally grown oxide, or TGO, which forms as the aluminum in the bond coat oxidizes. The third layer, a thin (around half a millimeter) ceramic top coat, has a low thermal conductivity and, therefore, acts as a barrier against heat damage.
"By applying a thermal barrier coating to a turbine blade, it is possible to increase the combustion temperature of the engine, which leads to significantly improved efficiency in gas turbines," said Dan Sordelet, an Ames Laboratory senior scientist. He explained that the ability of the bond coat to oxidize and form a continuous, slow-growing and adherent TGO layer is critical to creating a resilient and reliable thermal barrier coating.
Sordelet emphasized that cracking or breaking apart of the TGO layer due to time and service in a severe environment is one of the main causes of failure in a TBC system and the associated engine components. Also, at temperatures around 1100 degrees Celsius (2012 degrees Fahrenheit) and above, the aluminum in the bond coat begins to diffuse into the substrate, changing the overall bond coat composition.
"If enough aluminum diffuses into the substrate, eventually a phase change, which is a change in the crystal structure, occurs and can lead to large-scale distortion of the bond coat surface and subsequent failure of the TBC system," said Sordelet. Elaborating, he added, "Initially, there is a very thin TGO layer sitting on a very flat bond coat surface. If the bond coat continues to lose aluminum so that phase transformations take place, conditions will change from thin and flat to thin and 'rumpled.' Stresses develop, and the likelihood for the top coat to come off increases rapidly."
Working to improve the reliability of TBC systems, Sordelet and Brian Gleeson, director of Ames Laboratory's Materials and Engineering Physics Program and an ISU professor of materials science and engineering, have performed experiments on various nickel-aluminum-platinum, or Ni-Al-Pt, alloy samples made by Ames Laboratory's world-renowned Materials Preparation Center.
"Dan and I received funding from the Office of Naval Research to conduct fundamental research on the Ni-Al-Pt system, including experimental determination of isothermal phase diagrams," said Gleeson. "The phase diagrams provided much-needed guidance for elucidating the relationships between phase constitution/composition and properties in this system."
Quite unexpectedly, the two researchers found that platinum additions significantly improved the oxidation resistance of nickel-rich bulk alloys having the same type of structure as the turbine alloy. Without platinum, these alloys form a relatively fast-growing TGO scale that is prone to spall, or break up, during thermal cycling. By adding platinum, the alloys become highly resistant to oxidation, forming a tenacious, slow-growing TGO scale. But Sordelet and Gleeson weren't satisfied yet.
"In the typical design of alloys for oxidation resistance, you always find that adding a little sprinkle of this and a little sprinkle of that can have dramatic effects," said Sordelet. "Well, Brian's intuition to sprinkle either zirconium or hafnium was remarkably accurate." As the researchers added a little bit of either or both to the nickel-rich compositions, things improved tremendously.
"With the addition of hafnium, oxidation rates went down by up to an order of magnitude," Sordelet said. "We now have growth rates that are the lowest ever reported. It's quite remarkable!"
In current aluminum-rich bond coat alloys, only a very small amount (e.g., <0.1 wt.%) of zirconium or hafnium may be added to improve oxidation before adding too much is detrimental, causing catastrophic oxidation failure. In commercial coating production, it is extremely difficult to achieve an adequately uniform distribution of such a small amount of metals like these in a cost-effective way. "Fortunately, in the new nickel-rich bond coat, we have observed significant reductions in oxidation rates over a wide concentration, from 0.5 to 4 wt.% hafnium," Gleeson emphasized. "These are no longer 'trace' levels to a processing engineer and can thus be easily alloyed homogeneously throughout the material." This attribute gives Sordelet's and Gleeson's new coating a huge processing window, which they both say has been very desirable to people they've visited with in the coatings industry.
Their work with the bulk alloys led Gleeson and Sordelet to yet another fortunate result. They discovered that platinum changed the diffusion behavior of aluminum in their nickel-rich compositions. "Instead of aluminum going from the bond coat down into the substrate, it was moving up from the substrate into the bond coat," explained Gleeson. "This phenomenon is referred to as 'uphill diffusion,' and it's a consequence of the strong chemical interaction between aluminum and platinum. With our new bond coating compositions, the substrate can act as a large reservoir for aluminum and hence maintain the protective growth of the oxide layer."
The two researchers have recently demonstrated that their new coatings can offer significant benefits over current state-of-the-art bond coatings used in advanced TBC systems. "We have been working with an aeroengine manufacturer, and the results to date have been extremely encouraging," said Sordelet.
Ames Laboratory is operated for the Department of Energy by Iowa State University. The Lab conducts research into various areas of national concern, including energy resources, high-speed computer design, environmental cleanup and restoration, and the synthesis and study of new materials.