Generally, nickel-based superalloy possesses a good mechanical strength and an oxidation resistance, which makes it a candidate of structural material for high-temperature applications.^[1] Nickel-based superalloy components such as heat engine and heat exchanger are very often exposed to aggressive environments, which are likely to contain oxygen and other reactants such as sulfide, chloride and sodium salt, and mechanical stresses simultaneously.^[2–5] It has been shown that a condensation layer of alkali metal salt which is generally considered to be a mixture of the salts of Na₂SO₄ and NaCl is often observed on the surface of the superalloy. The existence of such a condensation layer on the superalloy surface could accelerate the oxidation which can considerably damage the service of high-temperature components,^[6] which is known as hot corrosion. Studies^[7–12] have shown that the interaction between creep or fatigue and hot corrosion could determine the service life of a superalloy at high temperature.

As is well known, nickel-based superalloys are classified into wrought, cast and powder metallurgy (PM) alloys according to the manufacturing routes.^[13] For cast and wrought superalloys, especially directionally solidified superalloys, the hot corrosion behavior has been widely studied.^[14–18] Some research of the coatings for resistance to hot corrosion of cast superalloys has already made much progress.^[19–24] PM superalloys are used in key components for turbine parts of advanced aero-engines. The PM process has solved the problems of severe segregation, non-uniform microstructure, coarse grains and poor hot working property for the high alloying degree in wrought and cast superalloy.^[25,26] Spray forming, also known as spray atomization and deposition, is a solidification processing technology, which was recognized as a special powder metallurgy process (a new type of molding process based on the powder metallurgy).^[27–29] Spray-formed superalloy has a similar microstructure to the conventional PM superalloy, and it has some advantages of fine-scale microstructure, lower levels of oxygen and short procedure.^[30–32] Explorations for performances of the spray-formed materials are in progress.^[33,34]

In the present study, isothermal hot corrosion tests at service temperature are conducted to obtain the corrosion kinetics of spray-formed superalloy FGH100. Typically, FGH100 possesses 50%–60% γ′ precipitates and the precipitates are mainly composed of a moderate amount of the primary γ′ phase with a mean size of approximately 0.4 mm–1.2 mm, a relatively higher amount of the secondary γ′ phase with a mean size of approximately 0.2 mm–0.4 mm, and a marginal amount of the fine tertiary γ′ phase with a mean size of approximately 30 nm–70 nm. Carbides and carbonitrides are uniformly distributed at grain boundaries. The microstructural development of the corrosion scale and its phase constitutions of oxidation layers are examined. Specimens of the superalloy FGH100 of determined size are pre-coated with Na₂SO₄-NaCl and tested at 700 °C at atmosphere in a furnace. The evolutions of the corrosion scale and subscale microstructure are studied by using electronic balance, scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS) and x-ray diffraction (XRD).

FGH100 superalloy was prepared by vacuum induction melting and vacuum arc melting methods. Spray forming experiment was performed at University of Bremen on SK-II in Germany (using nitrogen as atomizing gas and then hot isostatic pressed (HIPed) at 1140 °C and 150 MPa for 3 h. Samples in this testing were taken from the steady-state regions of the ingot. The chemical compositions (mass percent, wt%) of the spray-formed FGH100 superalloy are listed in Table 1.

Table 1.

Table 1.

Table 1. Chemical compositions of FGH100 superalloy (in unit wt%). . Ni Co Cr W Mo Al Ti Nb Ta Zr C B Bal. 21.75 12.77 4.12 2.95 3.60 3.54 1.51 1.52 0.03 0.023 0.022

Table 1. Chemical compositions of FGH100 superalloy (in unit wt%). .

Hot corrosion samples each with a size of about 10 mm (in length) × 10 mm (in width) × 3 mm (in thickness) were first ground by SiC abrasive paper (1000 grid), then ultrasonically cleaned in acetone and ethanol and dried at 100 °C. The 25 wt%NaCl + 75 wt%Na₂SO₄ saturated aqueous solution was sprayed on each surface of different specimens and dried on heated metal plates. The quantity of salt compounds coated on the alloy samples was about 4.0±0.5 mg/cm². They were subjected to isothermal hot corrosion in a furnace at a temperature of about 700±1 °C for 1 h, 3 h, 5 h, 10 h, 20 h, 50 h, and 100 h respectively, with a supplement of salt per 10 h. Then samples were air cooled to room temperature and washed in boiling deionized water for 30 min to dissolve the remaining salt. After being dried at 80 °C for 1 h, the mass of specimens was measured. Three specimens were used to obtain the averaged mass change. One kind of corrosion kinetics at 700 °C was determined through the relationship between mass change per unit area and corrosion time. The other one is the relationship between thickness increase and corrosion time. SEM and EDS were employed to investigate the morphologies of cross-section and surface. Phase constitutions of corrosion layers were detected by XRD.

There are two temperature-dependent regimes of hot corrosion as observed by peaks in corrosion rate curve, which are more aggressive than air exposure alone. The two corrosion rate peaks occur at 900 °C–1000 °C and 600 °C–750 °C. The high-temperature hot corrosion which proceeds at 900 °C–1000 °C is associated with the melting point of Na₂SO₄ (Tm = 884 °C). This study focuses on the low-temperature hot corrosion (at 600 °C∼750 °C), in which molten Na₂SO₄ + NaCl eutectic (630 °C) forms that acts to lower the liquidus temperature. Because of the molten Na₂SO₄ + NaCl eutectic temperature (630 °C), the corrosion behavior appears in the liquid phase.^[35] Based on the experimental results and theories,^[13,36–38] the hot corrosion mechanism of the spray-formed superalloy FGH100 is gradually clear.

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. Gibbs free energies for (a) oxidation and (b) sulfuration of alloying elements in a temperature range from 0 °C to 1000 °C.

The participation levels of alloying elements in hot corrosion are different, which depends on the content of a certain element and the free energy of oxidation or sulfuration.^[39] The standard Gibbs free energies for oxidation and sulfuration are shown in Fig. 6, which are calculated by HSC Chemistry Software. The free energies of formation for Ni, Co, Cr Al, and Ti oxides are more negative than those of their corresponding sulfides.

In the beginning of the reaction with the molten salt, the nickel- and Co-rich γ-matrix appeared to be the first phase dissolved:

It is a general and common reaction. With the generations of MS and MO, the activity of O²⁻ near the reaction interface increases, so that the chemical reaction below occurs:

The resulting products can diffuse to the free surface of molten salt and decompose for the lower activity of O²⁻ there. The decomposition products MO form a loose oxidation layer without any protectiveness. The MS will be left in the interface. Because Cr is also a major element in γ-matrix, the formation of Cr₂O₃ is the next only to NiO and CoO. What is more, the Cr₂O₃ has the priority to combine with the O²⁻ through the following reaction

The reaction consumes amounts of O²⁻, which will not only inhibit the dissolutions of NiO and CoO but also accelerate the decomposition of and near the free surface. After that, the Cr taking part in the reaction would exist in the form of radical ions due to its high stability until it reaches saturation point in the liquid phase. For this reason, there will be sufficient time to form a Cr₂O₃ layer with continuous structure, which is the key to further preventing the layer from being corroded. The precipitation of Cr₂O₃, of course, has a tendency to be close to the outside surface for the gradient of O²⁻ activity. Although the relatively stable γ′ precipitates are secondary, it will suffer partial or incomplete dissolving, bringing Al and Ti into this corrosion process. As with the NiO and CoO, the oxides of Al and Ti would be deposited in the vicinity of the free surface after dissolving in superalloy-melt interface as a result of the role of the gradient. However, the oxide films of Al and Ti are also non-compact and non-protective. The constant consumption of in the process of corrosion suggests that the progress should stop if the supplement were little.

Although Na₂SO₄ is dominant, it is proven beyond doubt that the addition of even small amounts of NaCl in Na₂SO₄ is much more harmful than only either NaCl or Na₂SO₄.^[40] The oxidation layers containing Cr₂O₃ and TiO₂ are sensitive to Cl⁻ attack, leaving behind a porous structure. What is more, NaCl contamination can easily induce extensive spalling.^[39]

According to the negative Gibbs free energies from 0 °C to 1000 °C shown in Fig. 7, the spinels and perovskites are formed from NiO, CoO, Cr₂O₃, Al₂O₃, and TiO₂ spontaneously, thus providing theoretical bases for understanding the occurrences of these composite oxides. It is a process of self-diffusion and mutual-diffusion.

Fig. 7.

	Figure Option ViewDownloadNew Window
	Fig. 7. Gibbs free energies for combination reactions of these simple metal oxides from 0 °C to 1000 °C.

In the course of its development, the gradient of O²⁻ activity is an extremely critical factor, which gives rise to the “uphill” diffusion, interactive permeation among components and the formation of protective or non- protective oxide layers.^[15]

Atomic diffusion could conduce to variation of vacancy concentration, so that the appearance of a porous structure in the inner corrosion layers is because of the on-going vacancy clustering and growing with the increase of oxidation time. The crack between the outer and inner layers is left for the dissolving of sodium salt nearest to the inner layer.

As a summary of the above analysis, it is proposed that corrosion behavior of the spray-formed FGH100 could be divided into two parts: oxidation-sulfuration and transfer of oxides. The former is the foundation of the latter, and the latter determines the difference between hot corrosion and conventional oxidation. A close connection exists between the two. Ultimately, the oxidation-sulfuration process, as well as the concurrent evolution of the diffusion structure of the corrosion scale in low-temperature hot corrosion examined in this study, may be better appreciated in the framework of multi-component hot corrosion phenomenology on the basis of corrosion kinetics and compositional distributions for various components.

During hot corrosion, the sample mass and layer depth increase with time going by at service temperature. The outer compositions of hot-corroded superalloy FGH100 surface are composed of NiO, CoAl₂O₄, NiAl₂O₄, and NiCr₂O₄, while the inner compositions are mainly sulfides and the residual substrate. Hot corrosion of spray-formed superalloy FGH100 is confirmed to be a combination effect of oxidation-sulfidation and transfer of oxides.