Abstract
Pure silicide coatings and Y–Ce modified silicide coatings were prepared on Ti–6Al–4V alloy by pack cementation process. The coating structures as well as their wear behaviours were comparatively studied. The results showed that the pure silicide coating and Y–Ce modified silicide coating prepared at 1080°C for 4 h had similar coating structures, mainly consisting of a TiSi2 outer layer, a TiSi middle layer and a Ti5Si4 inner layer. However, the microhardness of the Y–Ce modified silicide coating was higher than that of the pure silicide coating. The wear tests demonstrated that the mass loss of the pure silicide coating was about two times larger than that of the Y–Ce modified silicide coating, indicating the beneficial effects of Y and Ce on the wear resistance of silicide coating.
Keywords
Introduction
TC4 is one of the most widely used alloys in aerospace and power industries due to its low density and relative good strength retention properties.1,2 However, the rather poor wear resistance, especially the poor antiadhesive wear resistance of this alloy, reduced its potential use in tribological applications such as lightweight bearings and/or rubbing parts in internal combustion engines.3–5 A traditional approach to promote the wear resistance of titanium alloys is preparing protective coatings on their surfaces.6–8 Pack cementation process was suitable for preparing protective coatings on TC4 alloy because it can provide metallic bonding between the coatings and the substrates.9,10 It is considered that a strong coating/substrate bonding is beneficial for the wear resistance by suppressing spallation during sliding. 11
Silicide coatings have relative good wear resistance due to their high microhardness. However, the brittleness seriously hindered their long term applications. 12 Rare earths, with refining and purifying effects, were widely used in surface engineering because they can effectively promote the wear resistance of the coatings.12–15 Li et al. studied the effects of Y2O3 on the wear resistance of Si–Al–Y codeposition coatings prepared on γ-TiAl alloy, and found that the addition of Y in the coatings can effectively promote their wear resistance because Y can lead to larger coating hardness. 12 Sharma found that adding CeO2 in Ni–WC composite coatings could significantly increase their abrasive wear resistance. 14 Zhu et al. found that CeO2 can lead to finer grain structure of chromium coatings. 15 Addition of two rare earth elements in the coating may take full advantage of their positive effects. 12
In the present study, both pure silicide coatings and Y–Ce modified silicide coatings were prepared on Ti–6Al–4V alloy by pack cementation process. The microhardness and wear behaviours of the base alloy, pure silicide coatings and Y–Ce modified silicide coatings were comparatively studied.
Experimental procedures
Commercially Ti–6Al–4V alloy was used as the substrates for preparing both pure silicide coatings and Y–Ce modified silicide coatings, which were cut into 10×10×5 mm coupons. The surfaces of each specimen were polished up to 1000# SiC paper and ultrasonically cleaned in an acetone bath before coating preparation processes and wear tests.
The coatings were prepared by an electric tube furnace. The pack mixtures composed of 10Si–5AlCl3.6H2O–85Al2O3 (wt-%) were used to prepare the pure silicide coatings, while the pack mixtures composed of 10Si–1·5Y2O3–1·5CeO2–5AlCl3.6H2O–82Al2O3 (wt-%) were used to prepare the Y–Ce modified silicide coatings. Si, Y2O3 and CeO2 powders were used as the donor sources, AlCl3.6H2O was used as the activator and Al2O3 was used as the filler. Each kind of the powders were weighed according to the ratio exactly and then tumbled by a ball mill until their particle sizes were less than 50 μm. Before pack, the substrates were buried in the pack powders in an alumina crucible, which was sealed with Al2O3 based silica sol binder. The deposition temperature was chosen as 1080°C and the furnace was heated to this temperature at a rate of 15°C min−1.
The wear tests were conducted on a HT-1000 friction wear testing machine by a ball on disc type tribometer under dry sliding in the air. SiC balls with a diameter of 4·76 mm were chosen as the counterparts. All wear the tests were conducted at a gyration radius of 4 mm, a normal load of 5·194 N and a sliding velocity of 224rev min−1 for 1 h. The mass losses of the specimens after wear tests were weighed by an analytical balance with an accuracy of 0·01 mg.
An HV-1000 microhardness tester was employed to measure the microhardness of the base alloy and the coatings with a Knoop indenter at a load of 0·49 N and loading time of 20 s. X-ray diffraction (XRD, Panalytical X‘Pert PRO) was employed to identify the constituent phases of the coatings. The microstructure and the morphologies of were analysed by a scanning electron microscopy (SEM, JSM-6360LV) equipped with an energy dispersive spectroscopy (EDS).
Results and discussion
Coating structure
Figure 1 shows the BSE images and XRD patterns of the pure silicide coating and Y–Ce modified silicide coating prepared at 1080°C for 4 h. It is seen from Fig. 1a and a′ that many protuberances located on the coating surfaces, which would cause fluctuation of the wear friction coefficients at the early sliding stage. From Fig. 2b and b′, it can be seen that the structures of the two coatings were similar, consisting of a thick outer layer, a thin middle layer and a thick inner layer. EDS analysis results of the constituent phases of the two coatings are listed in Table 1. The concentrations of Si in the outer layer of the pure silicide coating were about 59·8 and 63·4 at-% (points 1 and 2), while that of Si in the outer layer of the Y–Ce modified coating was about 66·2 at-% (point 5), according to the Ti–Si phase diagrams 16 and the XRD patterns conducted on the coating surfaces (Fig. 1c and c′), the outer layers of both coatings were TiSi2. However, it is obvious that the concentration of Si in the outer layer of the Y–Ce modified coating is higher than that of pure silicide coating, while concentration of Ti is the opposite. The middle layers of the two coatings could be TiSi because the concentration of Si in both layers was equal to that of Ti (points 3 and 6). 16 However, these middle layers were too thin to analyse using XRD method. The concentrations of Si in the two inner layers were respectively about 44·6 and 41·5 at-% (points 4 and 7), together with the Ti–Si phase diagrams 16 and the XRD analysis results (Fig. 1c and c′), both inner layers were determined to be Ti5Si4. The above characteristics of the coatings can offer a composition gradient between the coatings and substrates, which led to relatively strong coating/substrate bonding strength.17,18
BSE images and XRD patterns of pure silicide coating and Y–Ce modified silicide coating prepared at 1080°C for 4 h: a surface image; b cross-sectional image; c XRD patterns of pure silicide coating; a′ surface image; b′ cross-sectional image; c′ XRD patterns of Y–Ce modified silicide coating
Microhardness distribution profiles, friction coefficients and mass losses of base alloy, pure silicide coating and Y–Ce modified silicide coating after siliding for 1 h: a microhardness; b friction coefficients; c mass losses
EDS analysis results of phases marked by pluses with numbers 1–7 in Fig. 1
It is obvious that the Y–Ce modified coating was more compact and denser than the pure silicide coating as considerable holes or microholes can be observed in the upper parts of the pure silicide coating. Another difference between the Y–Ce modified coating and the pure silicide coating was their coating thickness. The pure silicide coating had a thickness about 35 μm, while the Y–Ce modified coating possessed a thickness about 50 μm. This suggested that the addition of Y2O3 and CeO2 in the pack can promote the coating growth rate significantly. It is assumed that active atoms Y and Ce would preferentially absorb on surfaces of the samples during the coating formation process due to their higher activity, which would produce active centres and correspondingly increase the adsorption of active Si atoms during the coating formation process. This would lead to higher chemical potential and diffusion rate of Si; as a result, thicker and denser coatings were obtained. A denser structure would be beneficial for the coating to endure spallation and wear damages during sliding.
Microhardness
Figure 2a presents the microhardness distribution profiles of the pure silicide coating and the Y–Ce modified silicide coating from the coating surfaces to the substrate. It can be seen that the hardness values of both pure silicide coating and Y–Ce modified silicide coating were far larger than those of the substrate. The hardness distribution of both coatings showed a gradual decrease from the coating surfaces to the substrate, and this is mainly attributed to the different concentration of Si in each layers, i.e. higher Si concentration would correspondingly cause higher hardness. Moreover, the hardness values of the Y–Ce modified silicide coating is even higher than the pure silicide coating, which is connected with from its higher Si concentration and denser coating structure, as has been discussed above. Formation of a harder coating is also thought to be able to improve its wear resistance. 12
Friction and wear behaviour
Figure 2b illuminates the variations of friction coefficient–time curves for the base alloy, pure silicide coating and the Y–Ce modified silicide coating. It is observed that all the friction coefficients exhibited rapid increase in the running-in period. In the stable sliding stage, the friction coefficient of the base alloy was the largest, while that of the Y–Ce modified silicide coating was the lowest. Thus, it can draw a conclusion that both pure silicide coating and Y–Ce modified silicide coating had obvious friction reduction effect under the sliding condition against SiC. Moreover, the addition of Y and Ce in the coating imposed positive effect on reducing the friction coefficient.
Figure 2c compares the variation characteristics of the mass losses of the base alloy, the pure silicide coating and the Y–Ce modified silicide coating after sliding for 1 h. It is noted that the mass losses of the pure silicide coating and the Y–Ce modified silicide coating were far lower than that of the base alloy. Moreover, mass loss of the pure silicide coating was about two times larger than that of the Y–Ce modified silicide coating, which clearly illustrated the beneficial effects of Y and Ce on improving the wear resistance of the coating. The reduction of the mass losses could be attributed to the enhancement in surface hardness, and this will be detailed discussed later.
The SEM images of the worn surfaces of the specimens with and without coatings are presented Fig. 3. Meanwhile, EDS analysis of the typical constituent phases of the worn surfaces was also carried out to offer auxiliary information for understanding the wear mechanisms, and the results are listed in Table 2. It is obvious that the wear tracks of both pure silicide coating and the Y–Ce modified silicide coating were narrower than that of the base alloy, which suggested that the coatings possessed much better wear resistance. Moreover, the worn surface of the Y–Ce modified silicide coating was narrower and smoother than the pure silicide coating.
Wear morphologies of a, a′ base alloy, b, b′ pure silicide coating and c, c′ Y–Ce modified silicide coating after sliding for 1 h
EDS analysis results of phases marked by pluses with numbers 8–13 in Fig. 3
It is seen from Fig. 3a and a′ that the counterpart had brought serious wear on the base alloy. The magnified image in Fig. 3a′ highlights that the worn surface of the base alloy was characterised by a lot of dispersed white grains, many grooves and adhesive craters, demonstrating the occurrence of plastic deformation on the worn surface. It is also noticed that abundant oxygen (about 61·9 at-%) was detected in the white grains (point 8), implying the occurrence of oxidation during sliding. The above observations suggested that the wear mechanisms of the base alloy could be adhesive wear, abrasive wear and oxidation wear.
The load imposed on the friction surface can be classified as normal force and tangential force. The normal force can cause the convex of SiC ball cutting into the surface of the samples, while the tangential force can lead to relative motion between the friction pairs. The lower hardness value and higher toughness of Ti–6Al–4V alloy made that serve plastic deformation prone to occur on the worn surface during sliding process, which would produce large friction energy and cause oxidation and serious adhesive wear because plastic deformation makes the adhesive wear increase.19,20 Under the friction force during sliding, the surface engaged in the friction cannot maintain its integrity and crush into small abrasive particles, which would act as grains and cause abrasive wear.
The wear tracks of the pure silicide coating and Y–Ce modified silicide coating were similar, as shown in Fig. 3b, b′, c and c′. In contrast with the base alloy, plastic deformation could be hardly observed on the worn surfaces of both coatings. The worn surfaces of the two coatings can be classified into two obvious sections: one section was the wear zones with considerable cracks. EDS analysis results (points 9 and 11) revealed that the oxygen concentration in these zones were about 68·2 at-%, implying the occurrence of oxidation. Another section was the bare coating zones with similar chemical compositions (points 10 and 12) to their corresponding coatings before wearing, which demonstrated that these zones of the coatings had not been worn yet. The appearance of the bare coating in the wear tracks could be caused by the irregularities of the coating surfaces. The above observations suggested that the wear mechanisms of both pure silicide coating and Y–Ce modified silicide coating could be set as serve delamination wear, fatigue wear and minor oxidation wear.
For the coatings with and without Y–Ce, their higher hardness values and fragility made it hard for the occurrence of plastic deformation on the worn zones, but easier for the initiation and propagation of the microcracks in the near surface of the worn zones, which would subsequently cause spallation. Thus, the main wear mechanisms of both coatings can be set as delamination wear and fatigue wear. Moreover, The friction energy would also caused oxidation on the worn surface, and the oxidation products were supposed to be mainly composed of SiO2, as revealed by EDS analysis shown in Table 2 (points 9 and 11). It has been reported that thin and flexible SiO2 film between the friction pairs could protect the worn surface from further damage because it could act as a lubricating cover and prevent the direct contact of the friction pairs, which would significantly lower the friction coefficient. 21
However, it seems that the worn surface of the Y–Ce modified silicide coating was much smoother and shallower than that of the pure silicide coating. Together with the mass losses bar chart shown in Fig. 2c, it is obvious that addition of Y and Ce in the coating imposed positive effects on its wear resistance. One important factor responsible for this could be the denser microstructure and higher hardness of the Y–Ce modified silicide coating. It is well accepted that a denser and hard surface could effectively lower the plastic deformation and suppress the generation and propagation of the cracks on the worn surface, which in turn resulted in less mass loss during sliding. Another important factor could be the promoting effect of Y and Ce on the formation of the oxide film during sliding. It has been reported that the RE elements such as Y and Ce in the coating could act as nucleation sites and promote the formation of the SiO2 film between the friction pairs because Y2O3 and CeO2 had much lower free formation energies. 22 Since flexible SiO2 film was lubricate and protective, addition of Y and Ce in the coating would thus lead to lower friction coefficient and smoother worn surface.
Conclusion
The pure silicide coating and the Y–Ce modified silicide coating prepared at 1080°C for 4 h had similar coating structures, mainly consisting of a TiSi2 outer layer, a TiSi middle layer and a Ti5Si4 inner layer. However, the Y–Ce silicide coating was much denser and thicker than the pure silicide coating.
The Y–Ce modified silicide coating possessed higher microhardness and lower wear rate compared to the base alloy and the pure silicide coating, demonstrating that addition of Y and Ce can effectively promote the wear resistance of the silicide coatings.
When sliding against SiC ball, the wear mechanisms of the Ti–6Al–4V base alloy can be concluded as adhesive wear, abrasive wear and oxidation wear; while the wear mechanisms of the pure silicide coating and Y–Ce modified silicide coating changed to delamination wear, fatigue wear and minor oxidation wear.
Footnotes
Acknowledgements
This work was financially supported by the Shaanxi Science and Technology Research and Development Program (No. 2013KJXX-08). The authors would like to thank the Corrosion and Protection Research Lab from Northwestern Polytechnical University for various supports.