CHARACTERIZATION OF VACUUM PLASMA SPRAYED COBALT-NICKEL-CHROMIUM-ALUMINUM-YTTRIUM COATINGS

This paper analyzes the influence of the plasma spray distance on the microstructure and the mechanical properties of the Co32Ni21Cr8Al0.5Y coatings deposited with the vacuum plasma spraying (VPS) procedure. The microstructure and the mechanical properties of the plasma spray coatings were determined by the interaction of the Ar/H2 plasma ions with the powder particles when the transfer of the speed and temperature of ions on the powder particles occurs. The effect of interaction directly depends on the time of the interaction between ions and powder particles, which is defined by the plasma spraying distance. The powder is deposited by the plasma gun F4 at three substrate distances: 270, 295 and 320 mm. The coating with the best structural and mechanical properties was tested on the oxidation in a furnace for heat treatment without protective atmosphere at 1100°C in a period of 240 hours. The morphology of the powder particles was examined on the SEM. The microstructure of the layers in the deposited condition was tested by light microscopy. The coating with the best mechanical properties was electrolytically etched with 10% oxalic acid solution H2C2O4x2H2O. The analysis of the microstructure of the etched coating was performed by light microscopy and on the SEM, before and after testing the coating on oxidation. The microstructural analysis of the deposited layers was performed in accordance with the ’Pratt-Whitney’ standard. The mechanical properties of the layers were assessed through the examination of microhardness by the HV0.3 method and through bond strength tensile testing.

This paper analyzes the influence of the plasma spray distance on the microstructure and the mechanical properties of the Co32Ni21Cr8Al0.5Ycoatings deposited with the vacuum plasma spraying (VPS) procedure.The microstructure and the mechanical properties of the plasma spray coatings were determined by the interaction of the Ar/H 2 plasma ions with the powder particles when the transfer of the speed and temperature of ions on the powder particles occurs.The effect of interaction directly depends on the time of the interaction between ions and powder particles, which is defined by the plasma spraying distance.The powder is deposited by the plasma gun F4 at three substrate distances: 270, 295 and 320 mm.The coating with the best structural and mechanical properties was tested on the oxidation in a furnace for heat treatment without protective atmosphere at 1100°C in a period of 240 hours.The morphology of the powder particles was examined on the SEM.The microstructure of the layers in the deposited condition was tested by light microscopy.The coating with the best mechanical properties was electrolytically etched with 10% oxalic acid solution H 2 C 2 O 4 x2H 2 O.The analysis of the microstructure of the etched coating was performed by light microscopy and on the SEM, before and after testing the coating on oxidation.The microstructural analysis of the deposited layers was performed in accordance with the 'Pratt-Whitney' standard.The mechanical properties of the layers were assessed through the examination of microhardness by the HV 0.3 method and through bond strength tensile testing.Introduction ystems of CoNiCrAlY coatings have been developed based on the systems of NiCrAl, FeCrAlY, NiCrAlY and CoCrAlY coatings (Mrdak, 2010, pp.5-16), (Mrdak, 2012, pp.182-201), (Driver, 2004), (Feuerstein, et al., 2008, pp.199-213).The CoNiCrAlY coatings are used in different applications for the protection of gas turbines against high temperature oxidation and hot corrosion.Since the properties and the behavior of coatings are closely related to the microstructure, it is necessary to examine the structure of coatings after deposition and oxidation at elevated temperatures (Gudmundsson, Jacobson, 1988, pp.207-217).In order to understand the performance of Co-based CoNiCrAlY coatings better, it is necessary to understand the role of each element in the coating.With the increase of the Al content, its effect in the coating increases.Tha Al content should be high enough to form and maintain the α -Al 2 O 3 oxide layer which prevents subsequent oxidation (Prescott, Graham, 1992, pp.233-254).For this type of coatings, a typical content of Al is 10 -12 wt%.Aluminum in CoNiCrAlY alloys, forms the β (Co, Ni) Al phase which serves as a reservoir for the renewal of the protective α -Al 2 O 3 oxide.Co-based alloys with Al produce the β -CoAl phase which improves the resistance of these alloys to sulphidation, they also produce the β -NiAl phase which improves the resistance of these alloys to high temperature oxidation.It is also certain that the addition of Nickel to a Co-Cr-Al alloy reduces the process of interaction between the coating and the superalloy (Tamarin, 2002).The presence of Y improves the bonding of the α-Al 2 O 3 oxide with the coating (Bose, 2007), (Brandl, Tamarin, et al., 1998, pp.10-15).Peng et al. found that the presence of Y prevents the forming of cavities at the interface with the substrate (Peng, et al., 2003(Peng, et al., , pp.2293(Peng, et al., -2306)).Moreover, the Y content is of crucial importance for the growth of the TGO oxide on the coating surface (Toscano, et al., 2006, pp.3906-3910).A high content of Y leads to a high rate of the TGO growth, which is unfavorable and harmful for the coating.
CoNiCrAlY coatings are often deposited by the plasma spraying process in the vacuum (VPS).The development of the VPS technology has led to a significant improvement of the quality in coatings in comparison with the coatings produced at the atmospheric pressure.The main difference is that the process is performed in the vacuum in the absence of air at low pressure under the conditions of a high level of cleanliness and with the use of the transferred arc for cleaning the surface of the substrate.In its deposited condition, the microstructure of the CoNiCrAlY coating consists of two phases, γ and β.The γ phase is a solid solution of Co, Ni and Cr.The β (Co, Ni) Al phase is formed from S the β -CoAl phase and the β -NiAl phase.The present β phase and its share in the structure are essential for the protection of CoNiCrAlY coatings.The service life of the CoNiCrAlY coating in oxidation conditions is directly related to the amount of the β-phase which occurs in a variety of morphologies associated with different degrees of cooling related to different sizes of powder particles during spraying (Poza, Grant, 2006, pp.2887-2896).The elongated morphology of the β phase within γ grains and small β grains located on the border between the γ grains were associated with rapid cooling of melted small powder particles.Larger β grains were associated with larger particles and slower cooling (Poza, Grant, 2006, pp.2887-2896).In the Co-based CoNiCrAlY alloy there is no γ' phase present (Tamarin, 2002).The reason for the absence of the γ' phase in this alloy has been explained by some researchers (Achar, et al., 2004, pp.272-283), (Czech, et al., 1995, pp.28-33), who claim that Co tends to decrease the γ' phase.The stability of the β (Co, Ni) Al phase is reduced at high temperatures due to the diffusion of Al.Cheruvu and Mobarra with associates (Cheruvu, et al., 2000, pp.50 -54), (Mobarra, et al., 2006(Mobarra, et al., , pp. 2202(Mobarra, et al., -2207) ) have found that, at high temperatures, Al from the β phase fills the oxide layer on the coating surface and takes Al out of the β phase.By exposing the CoNiCrAlY alloy to 1100°C the TGO zone with a protective α -Al 2 O 3 oxide layer is formed on the surface.In the zone near the protective α -Al 2 O 3 oxide layer, there is no β-(Ni, Co) Al phase because the surface layer is Al depleted (Nicholls, Bennett, 2000, pp.413-428).Only a small amount of Al remains in the regions rich in (Ni, Co) (Leea, 2005, pp.239 -242).In this area, there is the Al-depleted β -zone which is below the upper TGO oxide layer.The thickness of the depleted β -zone increases with a longer exposure of the alloy to high temperatures due to aluminum consumption and the growth of the TGO layer (Nicholls, Bennett, 2000, pp.413-428).During the oxidation, protective oxide cracks and peels off from the surface and the aluminum from inner coating layers diffuses to the surface and restores a protective surface oxide layer (Nicholls, Bennett, 2000, pp.413 -428), (Wang, et al., 2002, pp.70 -75), (Gurrappa, Sambasiva, 2006, pp.3016-3029).In the TGO zone, besides the α -Al 2 O 3 oxide, there are spinel compounds such as CoAl 2 O 3 and NiAl 2 O 3 or (Ni, Co)(Al, Cr) 2 O 4 (Tang, et al., 2004, pp.228-233).Aluminum depletion near the surface leads to the transformation of the β (Ni, Co) Al phase into the Υ'-Ni 3 Al phase.The extending of oxidation causes the growth of this area and the transformation of the Υ '-Ni 3 Al phase into the Υ-solid solution.As a result, the coating degrades (Jiang, et al., 2010(Jiang, et al., , pp.2316(Jiang, et al., -2322)), (Mobarra, et al., 2006(Mobarra, et al., , pp.2202(Mobarra, et al., -2207)).The oxidation of the Υ phase in the depleted β -zone occurs with a faster formation of a protective oxide shell.CoNiCrAlY coatings in the deposited condition have a high bond strength of 55 -62MPa and micro hardness of 558 ± 43 HV 0.3 for the average value of porosity of 4.2% (Material Product Data Sheet, 2011, DSMTS-0092.1, Sulzer Metco).In addition to good mechanical properties, the coatings have a low coefficient of friction of 0.85 -0.9 and are resistant to wear (Gudmundsson, Jacobso, 1988, pp.207-217).The recommendation of the powder manufacturer for the CoNiCrAlY coating operating temperature is ≤ 1050°C (Material Product Data Sheet, 2011, DSMTS-0092.1, Sulzer Metco).
The paper presents the results of experimental investigations of the impact of spray distances at low pressure on the mechanical properties and the microstructure of Co32Ni21Cr8Al0.5Ycoating layers.Three groups of samples were made with three different distances of plasma guns: 270, 295 and 320mm.The coating with the best properties was tested on oxidation in a heat treatment furnace without protective atmosphere at 1100°C for a period of 240 hours.The main aim of this study was to make Co32Ni21Cr8Al0.5Ycoating layers homologous and to apply them on aeronautical parts exposed to a combination of high temperature oxidation and hot corrosion.The microstructure and mechanical properties of the coating layers were analyzed and the coating with the best quality was selected.

Materials for testing and samples
The powder produced by the 'Sulcer Metko' (Sulzer Metco) company ,marked AMDRY 9951, was used for the experiment.The Co32Ni21Cr8Al0.5Ypowder was developed for the production of coatings used to protect the base metal from high temperature oxidation and hot corrosion at temperatures T ≤ 1050°C (Material Product Data Sheet, 2011, DSMTS-0092.1, Sulzer Metco).The metal powder was produced by the atomization of liquid melted Co32Ni21Cr8Al0.5Yalloy with the inert gas of Argon.The produced particles of a spherical shape have a good flow in the jet plasma.Figure 1 shows the scanning electron microphotography (SEM) of the morphology of Co32Ni21Cr8Al0.5Ypowder particles.The range of the granulation of powder particles used in the experiment was from 5 to 37 µm.
The basis for deposited coatings for testing microhardness and evaluating the microstructure in the deposited condition was made of Č.4171 (X15Cr13 EN10027) steel in thermally unprocessed condition with the dimensions: 70x20x1.5mm(Turbojet Engine -Standard Practices Manual (PN 582005), 2002, Pratt & Whitney, East Hartford, USA).The samples for testing the coating microstructure on oxidation at 1100°C were made of alloy NIMONIC 80A with the dimensions: 70x20x1.5mm(Turbojet Engine -Standard Practices Manual (PN 582005), 2002, Pratt & Whitney, East Hartford, USA).

Examination of microhardness, bond strength, and microstructure
The evaluation of the mechanical properties of layers was done by examining the layer microhardness with the HV 0.3 method and by examining the bond strength by tensile testing.The microhardness was measured along lamellae, in the middle and at the ends of the samples.Five readings were performed and their values averaged.
The method of testing bond strength is the tensile testing method.The testing was done at room temperature with a tensile speed of 1cm/60s.Three test tubes were tested for each group of samples.
The morphology of powder particles was examined by the SEM method.The microstructure of layers in the deposited condition, after etching, was examined by light microscopy.The coating with the best mechanical properties, thermally treated to the oxidation at 1100°C for a period of 240 hours, was tested with a scanning electron microscope (SEM).The etching of the coating was done electrolytically with 10% of oxalic acid -H 2 C 2 O 4 2H 2 O.

Powder deposition
The powder was deposited at low pressure in the vacuum with a mixture of plasma gases Ar-H 2 .Figure 2 shows the vacuum plasma spray (VPS) system by the company 'Plasma Technik AG', designed for the protection of aeronautical parts exposed to a combination of excessive oxidation and hot corrosion.In the vacuum chamber there is a rotary table, a planetary system with 48 tools, a six-axis robot and an artificial arm.The handling system is designed in such a way that the tool and the substrates simultaneously rotate around their axes.Such complex movement allows even cleaning with the transferred arc and depositing powder evenly on the entire surface of the substrate.Table 1 shows the VPS parameters of the deposition of the Co32Ni21Cr8Al0.5Ypowder on the samples.The deposition of the Co32Ni21Cr8Al0.5Ycoating on the substrates of the samples was performed in the following way.The substrates were mounted into the supporting tools that were on the planetary system rotating around its axis.After the substrate mounting, the vacuum chamber was closed.The entire system is automated and programmed on the robot microprocessor unit.All parameters are given In the program.The process of vacuuming the chamber, the flow of plasma gas, cleaning of the substrate, the flow of powder, the deposition, the cooling of the substrate and the ventilation of the vacuum chamber are completely time-synchronized.In the sealed chamber, the artificial hand that accepts the tool with the substrate from the planetary system and sets it on the rotary table is on the other side of the chamber opening and cannot be seen in Figure 2.After the mounting of the tool with the substrate, the chamber was vacuumised and a pressure of 10 -3 mbar was reached in 5 minutes.Ar was then injected into the vacuum chamber through the plasma gun anode to the level of pressure of 25 mbar.At this pressure, all surfaces of the substrate were cleaned using the transferred arc.The distance of the plasma guns from the the surface of all the substrates was 270 mm.The plasma gun was set on (+) pole, and the substrate on (-) pole.This relation, called direct polarity, allows the oriented ions of the secondary gas He to clean the substrate surface from the impurities with a high speed and energy, making the substrate surface reactive.After the substrate cleaning, the powder was deposited on the substrates.The secondary plasma gas H 2 was added to the primary gas Ar.The pressure in the chamber was increased to the level of the operating pressure of 120 mbar.The constant pressure during deposition is provided by the vacuum pump.When the working pressure of 120 mbar was reached, the powder was injected into the plasma gun.The deposition rate is constant and does not change during the deposition.A layer of 0.1 mm is deposited in approximately one minute.When the deposition process was completed, the substrate was cooled in the chamber at a temperature of 300°C with Argon which flows from the plasma gun anode opening.The cooled substrate with the tool is taken by the artificial arm and returned to its original position.The planetary system turns for one step, so that the artificial hand can accept another tool with the substrate.The cycle of the powder deposition was repeated until the powder was not deposited on all the substrates.In this study, three groups of samples were made, with three distances of the powder deposition: 270, 295 and 320 mm.The coatings with thicknesses of 0.15 to 0.20 mm were formed.The other parameters were constant.The coating with the best structural and mechanical properties was tested on oxidation in the heat treatment furnace without protective atmosphere at 1100°C for a period of 240 hours.

Results and discussion
The measured values of the microhardness and the bond strength for deposited Co32Ni21Cr8Al0.5Ycoatings depending on the plasma spray distance from the substrate are shown in Figures 3 and 4. The values of layer microhardness are directly related to the distance of the powder deposition.The plasma spray distances in the vacuum significantly influenced the values of microhardness and bond strength of the deposited layers.The highest value of microhardness of 615 HV 0.3 was found in the layers deposited on the substrate with the lowest plasma spraying distance of 270 mm and with the lowest proportion of pores.The coating layers deposited with the highest distance of 320 mm had the lowest value of microhardness -490 HV 0.3 .The large distance from the substrate influenced the speed reduction and the subcooling of melted powder particles.The result of a larger distance is reduced sagging of one particle to another and the formation of pores throughout the coating layers.These values were confirmed by the analysis of the coating microstructure by using light microscopy.
The comparison of the values of tensile bond strength,showed that good values of the bond strength were obtained for all three plasma spray distances.The cleaning of the substrate surface by the transferred arc resulted in better adhesion of the deposited coating layers, which then resulted in obtaining higher values of the bond strength.The bond strength of the coatings significantly depended on the plasma spraying distance.A lower value of the tensile bond strength of 52MPa of the coating deposited with the highest plasma spraying distance of 320 mm resulted in a lower degree of fusion of powder particles in comparison with other two deposited layers.The highest value of bond strength of 78MPa was found in the layers deposited with the shortest plasma spraying distance.These layers were the thickest.The tensile testing of the bond strength showed that in all deposited coatings, the mechanism of failure took place at the interface between the substrate and the coating.Since the proportion of pores and unmelted particles is directly related to the values of the bond strength of the coatings, these measured values for the deposited coating with the lowest plasma spraying distance indicate that their share is the lowest in comparison with two other coatings.These values were confirmed by the analysis of the microstructure of the coatings by using light microscopy.For all the deposited coating layers, the mechanism of failure was adhesion at the interface between the substrate and the coating.
Figures 5, 6 and 7 show the microstructures of the deposited layers on the substrates with a plasma spray distances of 270, 295 and 320 mm.The coating microstructures are in non-etched condition.The qualitative analysis showed that there are no defects on the interface between the substrate and the deposited coating such as discontinuity of the deposited layers on the substrate, microcracks, macrocracks and separation of the coating from the substrate.Figure 5 shows the layers of the deposited Co32Ni21Cr8Al0.5Ycoating with the best structural and mechanical properties.The coating is thick and micropores cannot be observed through the layers, which is not the case with two other coatings.These layers are deposited onto the substrate with a plasma spraying distance of 270 mm continuously without interruption and without the presence of microcracks.Unmelted particles and precipitates are not present in the layers.The coatings deposited with a higher plasma spray distance have the presence of micropores of spherical and lamellar forms in their structure.The microstructures of the Co32Ni21Cr8Al0.5Ycoating layers shown in Figures 6 and 7 show spherical and inter-lamellar pores marked with black arrows.These layers were deposited with a higher substrate distances of 295 mm and 320 mm from the plasma gun.Due to larger distances of the plasma gun, there was the subcooling of melted powder particles, which were, on impact with the substrate, less plastically deformed by forming spherical and inter-lamellar pores of black color.There are no unmelted powder particles and microcracks in the structure.Through all layers of the deposited coatings,oxide lamella cannot be noticed since the VPS -vacuum plasma spray process allows depositing of layers without the content of oxides in the coating, which is a huge advantage when compared to the APS -atmospheric plasma spray process.The largest proportion of spherical and inter-lamellar pores was found in the Co32Ni21Cr8Al0.5Ycoating layers deposited with the highest plasma spraying distance of 320 mm.Because of the highest content of pores, these layers had the minimum value of microhardness and bond strength.The microstructures of the Co32Ni21Cr8Al0.5Ycoating deposited with a plasma spraying distance of 270 mm in etched condition with the best mechanical and structural characteristics, obtained by light microscopy, are shown in Figures 8 and 9.In the microstructure of the coating there are two phases γ + β which differ in color (Poza, Grant, 2006, pp.2887-2896), (Achar, et al., 2004, pp.272-283).The γ phase is light gray and the β phase is dark gray.The distribution of the phases in the microstructure is better seen in an SEM microphotography in Figure 10 where the coating is deposited with a plasma spraying distance of 270 mm.(Poza, Grant, 2006, pp.2887-2896), (Achar, et al., 2004, pp.272-283), (Czech, et al., 1995, pp.28-33).Unmelted powder particles and precipitates are not present in the structure of the coating, which indicates rather uniform distribution of powder particles in the vacuum with respect to the atmospheric pressure.
Figure 11 shows the (SEM) microstructure of the coating tested on oxidation in a heat treatment furnace without protective atmosphere at 1100°C for a period of 240 hours.The anticipated changes occurred in the microstructure of the Co32Ni21Cr8Al0.5Ycoating in comparison with the coating microstructure in the deposited condition.The exposition of the Co32Ni21Cr8Al0.5Ycoating to 1100°C decreased the stability of the β (Co, Ni) Al phase because of Al diffusion.Aluminium from the β phase filled the oxide layer on the coating surface and increased the thickness of the Al-depleted β -zone (light gray) marked with black arrows in Fig. 11 (Cheruvu, et al., 2000, pp.50-54), (Mobarra, et al., 2006(Mobarra, et al., , pp.2202(Mobarra, et al., -2207)).Besides α-Al 2 O 3 oxide, the TGO zone was formed on the coating surface.In the TGO zone, besides the protective α -Al 2 O 3 oxide, there are spinel compounds such as CoAl 2 O 3 and NiAl 2 O 3 or (Ni, Co) (Al, Cr) 2 O 4 (Tang, et al., 2004, pp.228-233).In the zone near the protective α -Al 2 O 3 oxide layer and the TGO zone, there is no β (Ni, Co) Al phase, because the surface layer is Aldepleted (Nicholls, Bennett, 2000, pp.413-428).Only a small quantity of Al remained in the regions rich in (Ni, Co) (Leea, 2005, pp.239-242).In the area below the TGO zone there is the Al-depleted β -zone.The thickness of the depleted β -zone is from 5 to 8μm, because of the long exposure of the coating to high temperature and due to the consumption of aluminum and the growth the TGO layer (Nicholls, Bennett, 2000, pp.413-428).In the lower layers of the coating, the untransformed light gray phase of the solid solution of γ -(Co, Ni and Cr) and the dark gray β (Co,Ni) Al phase are clearly evident.This indicates that the Co32Ni21Cr8Al0.5Ycoating did not degrade during 240 hours.

Conclusion
In this paper, the vacuum plasma spraying (VPS) process was used to deposit Co32Ni21Cr8Al0.5Ycoatings from the plasma gun at three distances of 270, 295 and 320 mm from the substrate.The structure and mechanical properties of the coatings were studied and analyzed in deposited condition as well as the influence of oxidation at 1100°C for a period of 240 hours on the microstructure of the deposited layers with the best characteristics, which led to the following conclusions.
The mechanical properties of the hardness and the bond strength of the coatings were directly related to the distance between the substrate and the plasma gun.Smaller distances of the substrate (270 mm) from the plasma gun gave the layers of coatings with higher microhardness and bond strength.Larger distances caused a formation of spherical and inter-lamellar pores through the coating layers; these pores resulted in lower values of the microhardness and the bond strength of these coatings.The mechanical properties of the coatings were in correlation with their microstructures.
The best microstructures were found in the layers deposited at a distance of 270 mm.These layers are dense and without the presence of micropores.The microstructure of the Co32Ni21Cr8Al0.5Ylayers of all deposited coatings is two-phase and consists of γ + β phases.The structure of the Co32Ni21Cr8Al0.5Ycoatings consists of the basic γ -(Co, Ni, Cr) solid solution (light gray), with the uniformly distributed β (Co, Ni) Al phase (dark gray).
(Co, Ni) Al phase rich in Al becomes Al-depleted at the coating surface.Besides the α-Al 2 O 3 oxide, the TGO zone is also formed at the coating surface.Besides the protective α -Al 2 O 3 oxide, the TGO zone contains spinel compounds such as CoAl 2 O 3 and NiAl 2 O 3 or (Ni, Co) (Al, Cr) 2 O 4 as a result of the diffusion of Co, Ni and Cr from the Y -solid solution.Below the TGO zone there is the Al-depleted β -zone in light gray.Lower coating layers still contain the untransformed γ -(Co, Ni and Cr) phase of the solid solution and the β (Co, Ni) Al phase.This indicates that, after 240 hours, the Co32Ni21Cr8Al0.5Ycoating proved to be resistant to high temperature oxidation at 1100°C for a period of 240 hours since it did not degrade.