High-temperature oxidation and hot corrosion behavior of the Inconel 738LC coating with and without Al₂O₃-CNTs

1 Introduction

Gas power plants in Iraq represent an important resource of electrical energy. To ensure the highest performance and efficiency in these plants, numerous components and units must, of necessity, correctly work together. The gas turbine is the principal unit, the very heart as it were, containing a large variety of highly advanced materials called superalloys, which can resist extreme thermal and chemical environments [1]. The nickel-based superalloys display face-centered cubic (fcc) solid solutions (γ) with sufficient quantities of Al and Ti, to produce the Ni₃(Ti, Al), fcc γ′ precipitate, and M₂₃C₆/M₆C carbides that are present at the grain boundaries [2,3]. Turbine blade failure stops the turbines from working, producing a power plant outage that can stretch from 1 to 4 weeks or even longer at times, depending upon the amount of damage done and the technique adopted to restore the machine back to normal running, all causing severe financial losses [4]. A few of the causes that lead to failure, either rapidly or slowly, are hot exhaust gases, high pressure, high speed, and the presence of heavy fuel contaminants. The most dominant reasons for failure induced by the high temperatures and presence of dangerous pollutants include hot corrosion (HC) and erosion corrosion [5]. For HC to occur no electrolyte environment is required; it takes place because of the deposition of impurities and salts like sodium sulfate and vanadium oxides [6].

Vanadium, an impurity present in fuel oil, can cause huge corrosion problems due to the formalization of V₂O₅, the melting point of which is 670°C [7,8,9]. High-temperature coatings can raise the corrosion resistance of Ni-based superalloys. Hence, the NiBSs find extensive use in gas turbine blades [10,11]. In fact, TBC is a kind of ceramic coating that is spread onto the constituents or elements that comprise a system to ensure the highest protection possible, against the most intense and extreme operating temperatures. A TBC, therefore, is composed of a metal bond coat, a metal substrate, and a ceramic top coat. Under harsh and severely intense settings of working conditions, the intermediate bond coat (MCrAlY) plays a pivotal part in the adherence of the ceramic top coat and thus raises the resistance to attack, normally induced by the molten salt and oxygen [12,13,14]. As they are well known for their remarkable mechanical properties, carbon nanotubes (CNTs) are widely used to create aluminum-based composites; the CNTs have been found to function as bridges across the voids and cracks and are distributed uniformly in the aluminum matrix [15]. Due to these good qualities, the coatings are deposited via the plasma spraying technique onto most materials that act as the substrate and which possess a similar melting point, such as metal alloys, ceramics, and cermet. Indeed, the TBCs sprayed with plasma have low thermal conductivity values and exceptional chemical stability, even when the temperatures increase to high levels [16,17,18].

The thermal coating of Al₂O₃ and its effects on the electrochemical and mechanical properties had been studied and reviewed by many authors. Thermal coating of Al₂O₃ has many applications in different fields of industry [19,20,21,22,23]. The novelty of the current study can be summarized by the following statements. Due to their exceptional mechanical properties of CNTs are used to fabricate aluminum-based composites. Also, CNTs form bridges across the voids and cracks when uniformly dispersed in the aluminum matrix. The uniform dispersion of CNTs in the aluminum matrix led to improve mechanical properties such as mechanical strength, thermal conductance, and electrical conductivity. CNTs are the strongest materials ever discovered with tensile strength reported to be 50 times that of steel and also have conductivity 1,000 times that of copper.

2 Experimental work

Specimens were collected by cutting the turbine blade made of Inconel 738LC to the specific dimensions of 20 mm × 15 mm × 3.5 mm of length, width, and thickness, respectively. Once these specimens were cut, they were ground with SiC grinding papers. Next, alumina paste was used to polish them, the chemical composition of the specimens is given in Table 1, and all tests were conducted according to ASTM standards.

As displayed in Table 2, the specimens were coated with an intermediate coating layer termed the bond layer, which has a thickness of 150 μm. Next, the specimens were coated with an outer ceramic layer made up of alumina having different weight percentages from the CNTs.

The plasma spraying technique involved the use of a special spraying unit (semi-automated), and the coating was accomplished by atmospheric plasma. The coating materials were combined in a ball mill once the bond layer coating was sprayed on the specimens whose surfaces were prepared before this step. The fuel contaminations were prepared by mixing 67 wt% of Na₂SO₄ and V₂O₅. The HC weight change tests were then performed at 650, 750, 850, and 950°C, prior to and after post-coating the specimens with the contamination, by ash (67 wt% Na₂SO₄ and V₂O₅).

3 Results and discussion

From Figure 1, the SEM of the coated specimens is shown; with (a) 100% Al₂O₃, (b) 96 wt% Al₂O₃–4 wt% CNTs, (c) 94 wt% Al₂O₃–6 wt% CNTs, and (d) 92 wt%; Al₂O₃–8 wt% CNTs. These images show that the coating layers are relatively high in porosity; the different coating layers show differences based on the degree of porosity. Microcracks may be present, initiated by the thermal stresses produced during the coating process. When porosity and microcracks are present, they prove to be detrimental to the coating properties. The SEM micrographs clearly showed partially melted and nonmelted particles or particle clusters, in almost all the coating layers.

Figure 1

SEM showed that the specimens coated with (a) 100 wt% A₂O₃, (b) 96 wt% Al₂O₃–4 wt% CNTs, (c) 94% Al₂O₃–6 wt% CNTs, and (d) 92 wt% Al₂O₃–8 wt% CNTs.

Agglomeration of the CNT particles or other particles in the nanoscale may be the cause for the poor or partial melting. Voids may be produced during the plasma spray process, as indicated through the SEM images. These voids result from a sequence of the plasma torch moving back and forth. They may be highly detrimental to the corrosion protection behavior as they could open a path for the passage of oxygen into the substrate. This action was used to analyze the SEM images and thus assess the porosity percentage of the coating layer. During the plasma spraying process, the pores and micro-cracks can arise from a variety of sources, for instance, the entrapped gases, incomplete filling in the splats, which showed quick solidification, the shrinking of the splats during the latter process, and so on. The porosity of the plasma sprayed coatings can range from below 2% to an excess of 20%, depending on the coating process parameters, as well as the properties of the coating powders utilized. Specimens coated with (b) 96 wt% Al₂O₃–4 wt% CNTs, (c) 94 wt% Al₂O₃–6 wt% CNTs, and (d) 92 wt% Al₂O₃–8 wt% CNTs possess a relatively lesser degree of porosity. The splats are smaller in size and irregular in shape. These images indicate that the coatings have displayed a less degree of porosity, as well as fewer microcracks.

Measurements of the roughness of the surface coating layers are shown in Table 3. The C-coated specimen showed the highest roughness values, while the B specimen revealed the lowest surface roughness values, by AFM. Porosity induced by the plasma spray coating revealed a relatively high degree of porosity due to this technique, which subsequently tends to raise the surface roughness. The roughness of a surface can have a significant part to play in its corrosion performance. Agglomeration, inevitable in nanoscale powders, could result also in raised surface roughness of the coated specimen. This result corresponds to that of the researcher [24]. In Figure 2, the three-dimensional images and the grain histogram are revealed by the AFM test. The average roughness enables the sharpness of the surface of the film to be described; this is defined as the mean height of the surface from the center of the plane, in a given area.

Figure 2

AFM showed the specimens coated with: (a) 100 wt% A₂O₃, (b) 96 wt% Al₂O₃–4 wt% CNTs, (c) 94 wt% Al₂O₃–6 wt% CNTs, and (d) 92 wt% Al₂O₃–8 wt% CNTs.

The coat is used to study the actions of the specimen that had undergone continuous changes in weight, during the test intervals. From Figure 3, the weight change per unit surface area is evident as a function of time and expressed in the number of cycles for the bare specimens. The curve shows clearly that the maximum weight gain occurs at 950℃, which is possibly due to the high oxidation rate at this temperature. With the increase in time, the thickness of the oxide layer also increased, i.e., the stresses in the layer increased because of the cracks and voids that were present.

Figure 3

Bare base alloy oxidation behavior at 950°C temperature at different times.

In Figure 4, the XRD results of the specimens exposed to oxidation reveal that 650, 750, 850, and 950℃ for 50 h are listed. The oxides formed include NiO, TiO₂, and spinel phase NiCr₂O₄. All these oxide types are the pores that do not adhere very well to the base metal surface, and hence, they are not the protective oxide layer. The Cr₂O₃ that was visible at the interface between the alloy matrix was a layer abundant in nickel oxide. When the Cr₂O₃ sub-layer was continuous, an outer NiO scale was formed, with an interior layer, which sometimes appeared porous in nature, composed of NiO, and included the NiCr₂O₄ islands. The Nio layer stopped developing and the kinetics of oxidation barely displayed the growth of the Cr₂O₃ layer. When the spinel NiCr₂O₄ was introduced into the external layer, it caused a chemical reaction between the Cr₂O₃ and NiO layers. Some NiCr₂O₄ was produced by the oxygen moving ahead of the NiO front. In all the Cr alloys, (except for the most dilute ones) these NiCr₂O₄ islands continued to remain as a second phase in the scale, and their degree revealed the site of the original metal surface.

Figure 4

XRD patterns of bare specimens at 850℃ after 50 h postoxidation: (a) T = 650°C, (b) T = 750°C, (c) T = 850°C, and (d) T = 950°C.

The HC test performed was based on the weight change; test results were obtained at different times with different temperatures; weight change is a good indicator to study the corrosion rates under equivalent conditions. Weight change is measured per the corresponding time cycle. Figure 5 indicates the HC behavior (weight change) in artificial ash having a composition of 67 wt% V₂O₅ and 33 wt% Na₂SO₄ at 650, 750, 850, and 950℃. At 950℃, the specimen displayed much higher rates of weight gain, which continued until the last cycle. The weight change at 650°C showed that the specimen had reduced drastically in comparison with the other temperature. The change in weight observed in the specimens, in the presence of artificial ash, had increased, when compared with those of oxidation, without artificial ash salts.

Figure 5

The relation between weight change and different temperatures at the present artificial ash, after 50 h for coated specimens.

Subsequently, the XRD analysis was performed to investigate the corrosion products present on the outer surface of the specimens. The XRD patterns of the HC test after subjecting the specimens to artificial ash (33 wt% Na₂SO₄ and 67 wt% V₂O₅) and a salt environment at 650, 750, 850, and 950℃ for 50 h are revealed in Figure 6. From the XRD analysis, the presence of NiO and Cr₂O₃ was detected. Just above the substrate, a thin layer of NiO had been formed; above this layer, a continuous thin layer of Cr₂O₃ was noted, which had been formed with the outer portion of the scale, and containing the oxides of Ti Ni Cr, NiCr₂O_4, and Ni (VO₃)₂. Spinels, recognized by XRD analysis, were also visible in the scale. The spinel formation of NiCr₂O₄ through the reactions of the solid phases between the Cr₂O₃ and NiO in the oxide scales facilitates the oxidation resistance to advance. This occurs because the spinel phase typically possesses lower diffusion coefficients of the anions and cations than those corresponding values in the parent oxides. The XRD analysis also suggests the presence of Na₂CrO₄, and this finding concurs with the results of the research [25,26,27,28].

Figure 6

XRD patterns of specimens at different temperatures after 50 h, with artificial ash for coated specimens: (a) T = 650°C, (b) T = 750°C, (c) T = 850°C, and (d) T = 950°C.

From Figure 7, the SEM of the coated specimens for those covered with artificial ash for 950°C at 50 h is evident. These images clearly reveal that the coating layer retained its topography and soundness, implying its stability as a coating layer and that it may be acting as a barrier layer between the outside and substrate, thus averting the passage of the corrosive species and oxygen to the substrate.

Figure 7

SEM micrograph of Al₂O₃–8 wt% CNTs for specimen tested after 50 h at 950°C.

4 Conclusions

Based on the results of tests, the following conclusions can be drawn out:

• The oxidation and HC rate correspondingly rise with the increase in temperature, and the substrate layer becomes porous and weakly adherent.

• From the results of the SEM, cracks are visible on the surface of the specimens.

• The coated specimens reveal good resistance to HC and the additional CNTs further enhance the properties of the coating layer.

• The XRD results for the coated specimens with artificial ash are noted in several phases like the NaV₃O₈, NaVO, V₂O₅, and Na₂SO₄.