27TP Hoar and EAG Croam J Iron Steel Inst Vol 169 1951 p 101

28. L.L. Seigle, Surface Engineering, R. Kossowsky and S.C. Singhal, Ed., Martinus Nijhoff 1984, p 349-369 Properties of Diffusion Coatings on Superalloys

Coating Formation Mechanisms. Diffusion aluminide coatings on superalloys are classified by microstructure as being of the "inward diffusion" or "outward diffusion" type according to the seminal work of Goward and Boone (Ref 29). The classification was derived from studies of aluminide coating formation on a typical nickel superalloy, Udimet 700, which has a nominal composition of Ni-15Cr-17Co-5Mo-4Al-3.5Ti. It was observed that for pack mixes containing pure aluminum (unit or "high" activity), coatings formed by predominant inward diffusion of aluminum through Ni2Al3, and deeper in the coating, through aluminum-rich NiAl (for pure nickel, by inward diffusion through Ni2Al3 only). The diffusion rates are abnormally high—practical coating thicknesses can be achieved in a few hours at 760 °C (1400 °F). A typical as-coated microstructure is shown in Fig. 3(a). Upon further heat treatment at, for example, 1080 °C (1975 °F) for four hours, the microstructure shown in Fig. 3(b) is formed—the coating matrix is now NiAl. The single-phase region in the center of the coating is nickel-rich NiAl grown by predominant outward diffusion of nickel from the substrate alloy to react with aluminum from the top layer. The inner layer, or so-called interdiffusion zone, consists of refractory metal (tungsten, molybdenum, tantalum, etc.) carbides and/or complex intermetallic phases in a NiAl and/or Ni3Al matrix, formed by the removal of nickel from the underlying alloy, thereby converting its Ni-Ni3Al structure to those phases. Conversely, if the activity of aluminum in the source is reduced by alloying with, for example, nickel or chromium, to a level where nickel-rich NiAl is formed at the surface, the coating, shown in Fig. 3(c), grows by predominant outward diffusion of nickel from the substrate to form NiAl by reaction with aluminum from the source. The lower layer of this coating is formed as previously described. Diffusion rates are relatively low so the coating process must be carried out at higher temperatures—usually greater than 1000 °C (1830 °F). These mechanisms are consistent with those observed by Janssen and Rieck (Ref 30) and later by Shankar and Seigle (Ref 31) during studies of diffusion in the simple nickel-aluminum system. Figure 4 shows the ratios of diffusion rates of nickel and aluminum across the range of stoichiometry of NiAl (Ref 31). At the high aluminum limit of NiAl, diffusion is by predominant motion of aluminum, confirming the earlier postulate of Goward and Boone (Ref 29). A coating with a matrix of NiAl formed by this diffusion mechanism is shown in Fig. 3(d). Upon further heat treatment, this coating will stabilize with a structure similar to that shown in Fig. 3(b).

Fig. 3 Archetypical microstructures of aluminide coatings on a nickel superalloy. (a) Inward diffusion based on

Ni2Al3 (and aluminum-rich NiAl). (b) Same as (a) but heat treated at 1080 °C (1975 °F). (c) Outward diffusion of nickel in nickel-rich NiAl. (d) Inward diffusion of aluminum in aluminum-rich NiAl. Source: Ref 29

Fig. 4 Ratio of diffusion coefficients of nickel and aluminum as a function of aluminum in NiAl. Source: Ref 31

The above mechanisms apply equally to those coatings formed by out-of-contact or CVD processes, from slurry "slip packs" (Ref 32), and from aluminum alloy powders deposited on superalloys by slurry spraying or by slurry electrophoresis (Ref 33). Coatings applied by spraying (Ref 34, 35) or electrophoretically depositing (Ref 33) pure aluminum or low-melting aluminum alloys, for example, Al-10Si, and then heat treating, form by dissolution of the superalloy into the melt until the melt solidifies, followed by diffusion of aluminum similar to that described above.

All known aluminide-based coatings on nickel superalloys, including those modified by chromium, platinum, and silicon, have one of the archetypical microstructures described above. For pure nickel and nickel alloys containing no aluminum, (or <0.2% Al), the interdiffusion zone does not form. For pure nickel, Kirkendall voids and alumina, from oxygen in the nickel, form at the coating/substrate interface (Ref 29). For nickel alloys containing no aluminum, voids, refractory metal layers, and alumina form at the interface (Ref 29). The adherence of the resulting coatings is compromised and they may not be practically useful. It is anticipated that similar mechanisms apply to the coating of cobalt superalloys. Again, the absence of aluminum in many of these alloys precludes the formation of the interdiffusion zone common to most nickel superalloys. Rather, a refractory metal (tungsten, chromium) carbide forms at the juncture to the base alloy (Ref 40). As described for similar nickel-base alloys, this refractory metal carbide and alumina formed from oxygen in the aluminum-free alloys, can also compromise the adherence of these coatings. Special processing conditions, involving slow coating growth at high temperatures (up to 1095 °C, or 2000 °F) from relatively low aluminum activity sources, can sometimes be used to achieve satisfactory coating adherence. Minor additions of aluminum (1 to 2%) to cobalt superalloys completely obviate these problems—stable interdiffusion zones then form analogous to those on most nickel superalloys (Ref 36).

Diffusion chromide coatings formed on a nickel superalloy by pack cementation and out-of-contact processes are illustrated in Fig. 5. The coating deposited by pack cementation is overlaid with a thin layer of alpha-chromium as shown in Fig. 5(a). Users generally require that this phase be absent. It must then be removed chemically, or alternatively the coating applied by an out-of-contact process to produce the structure shown in Fig. 5(b). These coatings usually then contain chromium to the extent of 20 to 25% at the outer surface. Coating formation, from chromium-alumina-activator (usually ammonium chloride) packs or from out-of-contact sources (powders or chromium granules as described in Ref 18) involves approximately equal rates of interdiffusion of chromium and nickel. Significant depletion of titanium and aluminum from the alloy surface occurs because the sources do not contain these elements. The desired coatings are thus solid solutions of chromium in the remaining nickel-base alloy. Internal oxides of aluminum and titanium can form because the oxygen potential of the sources is normally sufficient to cause internal oxidation. This can be avoided by adding aluminum to the sources in amounts just below that which would cause aluminizing rather than chromizing (Ref 38).

Fig. 5 Chromium diffusion coatings on a nickel superalloy by (a) pack cementation and (b) out-of-contact gasphase processing. Both at 500x. Source: Ref 37

Fig. 5 Chromium diffusion coatings on a nickel superalloy by (a) pack cementation and (b) out-of-contact gasphase processing. Both at 500x. Source: Ref 37

Rapp and co-workers (Ref 16) have refined the theory of codeposition of aluminum, chromium, silicon, and reactive elements (yttrium and hafnium) by pack cementation and related processes and have demonstrated the benefits derived therefrom. These processes and coatings should find practical applications in the near future.

Coating Protection and Degradation. Simple aluminide coatings resist high-temperature oxidation by the formation of protective layers of alumina and can be used up to about 1150 °C (2100 °F). The coatings degrade by loss of aluminum due to spalling of oxides under thermal cycling conditions. Incorporation of reactive elements, such as yttrium and hafnium, by codeposition during aluminizing (Ref 16) can significantly improve adherence of the protective alumina scales and therefore extend coating life. At temperatures above about 1000 °C (1830 °F) interdiffusion of the coatings with substrates contribute significantly to degradation (Ref 39). Practical coating service lives are limited to operating temperatures of 870 to 980 °C (1600 to 1800 °F) with only short excursions at the highest temperatures.

Chromium modifications, made by diffusion chromizing prior to aluminizing (Ref 14) or by codeposition of aluminum and chromium (Ref 16, 41, 42), have enhanced resistance to various forms of molten-salt hot corrosion. Electroplating with a thin layer of platinum (and possibly rhodium) followed by aluminizing (Ref 15) forms a coating with substantially improved resistance to both oxidation and high-temperature (Type I) molten-salt corrosion. Additions of up to about 5% Si improve both oxidation and hot corrosion resistance (Ref 13). Silicon can be codeposited with aluminum by pack cementation (Ref 16, 43) and related out-of-contact processes. So-called slurry processes wherein a liquid suspension of aluminum and silicon powders is applied to the alloy surface, then dried and fired at elevated temperatures, can also be used to incorporate silicon (Ref 35).

The oxidation and hot corrosion resistance of these coatings are more or less influenced by the composition of substrate alloys. Tantalum and hafnium improve cyclic oxidation and hot corrosion resistance, the latter element by improving the adherence of the protective layer of alumina (Ref 44). Molybdenum and tungsten compromise hot corrosion resistance.

Because of the brittle fracture behavior of NiAl up to temperatures of 650 to 775 °C (1200 to 1400 °F), all aluminide coatings exhibit such fracture below these temperatures while above these limits ductile behavior occurs (Ref 45). This behavior can either compromise or enhance thermal fatigue resistance of substrate alloys depending on such complex factors as the exact nature of the thermal cycle and the structure--equiaxed, directionally solidified, or single crystal--of the alloys (Ref 46). If these effects are limiting, designers may require use of more expensive overlay coatings of the MCrAlY (M = Co and/or Ni) and/or thermal barrier (zirconia) types.

0 0

Post a comment