Oxidation of MCrAlY coatings on Ni-based superalloys

Ensuring a sustainable power generation supply into the future is of worldwide importance. A combination of power generation methods to meet this (including both fossil plant and renewables) is a certainty. Therefore, the maintenance of conventional fossil-fired plants is required to provide a low cost, flexible and reliable electricity supply, and in order to conserve fuel resources it is important to ensure they operate as efficiently as possible. Lifetime predictions of important components (such as turbine blades in industrial gas turbine engines) are essential in order to avoid failures and unnecessary downtime, improve the ability to schedule routine maintenance work and to help develop new coating systems to meet the ever more demanding requirements. Better models would allow improved predictions and mean that service schedules could be set with better accuracy thereby, limiting downtime and reducing costs as well as being better able to predict coating performance and thus increase engine efficiency. Ni-based superalloys are a common material utilised for turbine blades due to their high temperature strength. Coupled with MCrAlY coating systems and thermal barrier coatings (TBC) they are able to withstand both the high temperatures and arduous environments they will encounter during service. In order to effectively predict the lifetime of components such as turbine blades it is vital to understand the oxidation behaviour of these systems and how factors such as the MCrAlY composition, processing factors and thermal cycling affect their performance. MCrAlY coatings are commonly utilised to protect the alloy substrate from the high risk of oxidation and corrosion. The M in MCrAlY is generally Ni, or Co or a combination of the both. These systems work by elements within the MCrAlY (notably Al) reacting with oxygen in the environment and forming protective thermally grown oxides (TGO). As these systems are very complex, a detailed understanding is needed of all areas, especially the TGO. This research was concerned with the characterisation of the TGOs formed on different compositions of MCrAlY coatings as a function of time and temperature. A range of coated samples were exposed to both isothermal and cyclic conditions to simulate service conditions. A wide range of analytical techniques have been utilised including scanning electron microscopy (SEM), energy dispersive x-ray analysis (EDS), electron backscattered diffraction (EBSD), transmission electron microscopy (TEM) and dual beam focussed ion beam field emission gun scanning electron microscope (FIBSEM) for phase characterisation. Detailed analyses of the MCrAlY bond coat, TGO and TBC layers have been made with regards to how composition, temperature, processing parameters and thermal cycling affect them. Thermodynamic equilibrium calculations have also been performed using commercial software in conjunction with a thermodynamic database for Ni-based superalloys to provide information about phase stabilities at equilibrium. Four MCrAlY coating systems have been analysed in detail in this research (NiCoCrAlY, NiCrAlY, CoCrAlY and CoNiCrAlY). Thermodynamic equilibrium calculations predicting phase amounts and compositions have been compared with experimental data with good correlation. The microstructural evolution from short term to long term ageing has been charted for all bond coat systems. The Co containing systems were shown to be very similar showing a main aluminium rich TGO, with precipitates forming in the TGO multi-oxide/spinel regions above it. The NiCrAlY was shown to have a completely different behaviour, forming a mixed Cr/Al oxide initially followed by an aluminium rich TGO. Whilst precipitates were seen in the aluminium TGO, no multi-oxide/spinel regions were seen. Further in depth analysis of the oxide formations was undertaken. The oxide on all coating systems was shown to form via similar mechanisms, whereby the Y in the bond coat is doping the oxide and changing the growth mechanism. Analysis of the TGO showed that on the Co containing coatings the TGO was alumina and on the NiCrAlY a mixture of chromia, mixed Al/Cr oxide and alumina beneath. The form Y has taken in the bond coat was quantified using high resolution techniques, as were the Y precipitates within the TGO. Each bond coat system was seen to contain different forms of precipitate however they were all thought to mainly form via the outward diffusion of Y from the bond coat, although it is possible some were existing particles in the bond coat enveloped by the growing oxide. It was also discovered that in the NiCoCrAlY system Ti precipitates were present along with strong Ti segregation at the boundaries, where Ti was thought to have diffused from the superalloy substrate to the oxide. The presence of Ti is thought likely to alter the grain boundary diffusion and probably increase scale growth. The multi-oxide/spinel regions (shown to be mixed Co, Cr, Al, Ni oxides) were shown to correlate well with similar formations noted in previous literature and are thought to be a result of the variable interface morphology of the coatings resulting in localised Al depletion and the formation of breakaway oxides. The effect that processing parameters and thermal cycling can have on the system has also been investigated. It was shown that the interface morphology (which is linked closely to the processing parameters) can greatly affect the oxide formation, both in terms of precipitates within the TGO and multi-oxide/spinel regions. The interface morphology was also shown to have important implications with regard to the cyclic life, through its effect on the stresses within the system, which in turn can affect the cyclic life. The nature of the oxide formations were also shown to influence the cyclic life.