Microstructure and Tribological Performance of HVAF-Sprayed Ti-6Al-4V Coatings

Abstract

Ti-6Al-4V is a widely used titanium alloy in aviation and bio/chemical applications for its attractive mechanical and corrosion resistance properties. The use of Ti-6Al-4V as a coating for repair purposes through thermal spray techniques provides a unique productivity opportunity. A repair coating must be dense to provide the required in-service functionalities, such as resistance to wear. The High Velocity Air Fuel (HVAF) thermal spray technique deposits dense coatings with reduced concern for oxide inclusions. This work presents an investigation of the microstructure, dry sliding, and solid particle erosive wear performance of four different coatings engineered through the configuration of the nozzle of an HVAF spray gun, based on the length of the nozzle and the size of the nozzle exit. A long nozzle length and wide nozzle exit mean increased inflight dwell time and reduced average inflight temperature for the sprayed particles, respectively—a reversed configuration means the opposite. The tested coatings showed a porosity of less than 2%. The sliding and erosion wear performance of the densest of the coatings compares to that of the bulk material tested under the same conditions. Electron microscopy was used to investigate the driving mechanisms for the performance of the respective coatings. The implications of the results are discussed for the potential adoption of HVAF-sprayed coatings in metal component repair.

1. Introduction

Ti64 alloy makes up 50% of the alloys of titanium used in biomedical, chemical, automotive, and mostly in the aerospace industry. The aerospace sector consumes 80% of Ti64 alloy, while other applications account for the remaining 20% [1]. Ti64 alloy is designed for a good balance of engineering properties like the following:

• Ductility and workability;
• Fracture toughness;
• Heat and creep resistance;
• Weldability, thermal processability, and strength [2].

Titanium materials are the usual feedstock for powder-based additive manufacturing and for new repair technologies, such as thermal spray techniques [3]. Titanium alloys are very useful, but they are costly. It is cost-effective and environmentally friendly to be able to repair and remanufacture damaged structural engineering components made of titanium alloys to extend their service lifespan—instead of complete replacement. The repair of worn/damaged engineering components forms part of the ongoing sustainability efforts in the aerospace industry to combat climate change.

Repairs are a major issue in the aerospace industry, and there are ongoing efforts to improve subsisting manufacturing methods, including repairs, to cut costs and reduce operation losses [4]. While many of the components of an aircraft engine are still being made from Ti64 alloys (see Figure 1), there is a need to find an economic and sustainable way to further explore the much-appreciated properties of the alloys. Repairs in aero engine parts are mostly performed by welding Tungsten Inert Gas (TIG); where welding is impractical, the parts are replaced [5]. Conventional welding processes engender an energetic impact on parts being repaired and can make the part heavier. Welding also creates heat-affected zones in the substrates to cause changes in the properties of the affected zone—the localized change in properties gives further concern [6,7]. Thermal spray is an attractive material deposition method to create coatings needed for the surfacing and resurfacing of parts [8]. Thermal spray offers many advantages, such as low heat transfer to the substrate, improved engineering performance, and extending the component’s lifespan at a relatively low cost. Thermal spray can deposit coatings of varying thicknesses, from a few microns to several millimetres. Thermal spray coatings have been sprayed from all classes of materials, including metals, ceramics, polymers, ceramics, and cermet, without concern for the form of the material being a powder, wire, rod, dispersed in a liquid (suspension), or formed in situ from a solution precursor [9,10,11,12,13].

From a repair point of view, thermal spray produces dense, homogenous coatings that add insignificant mass to the size of a component. Despite the insignificant size of a coating relative to a coated component, a dense homogenous coating offers similar protection, as will a bulk sample of the coating material feedstock [14]. The key to ensuring the controlled microstructure of a coating lies in the adopted spray technique [15], the spray parameters/conditions [16], and the characteristics of the feedstock materials [17]. The aerospace industry has explored this capability of thermal spray with the high-velocity oxy-fuel (HVOF) thermal spray technique for repair. The HVOF repair route shows significant merits over welding; it reduces part distortion after repair in addition to coating thin, porosity-free layers over the damaged parts. However, HVOF overlay coatings are rough, and a part coated with HVOF requires secondary machining steps to ensure an improved surface finish—this is besides the possibility of in-flight oxidation of the feedstock deposited via HVOF. This is a concern for oxidation-sensitive materials like titanium alloys, which have also limited their application. For this work, the feedstock is a Ti-6Al-4V powder deposited via HVAF, a thermal spray technique with the capacity to deposit dense coatings with reduced/no oxide formation while transferring limited heat to the coated substrate—HVAF coating also maintains good interfacial bonding with the coated substrate [2,17]. The earlier stated benefits guided the choice of HVAF above the known industrially employed techniques APS and HVOF [18,19]. This paper aims to investigate the implication of HVAF hardware spray configuration on the microstructure, mechanical properties, and wear resistance of HVAF-sprayed Ti-6Al-4V coatings deposited on Ti-6Al-4V substrates. Coatings were deposited using four different spray gun configurations; the respective coatings were then studied from a processing-property-performance perspective to elicit a pathway for developing coatings usable for repair purposes.

2. Experimental

2.1. Materials

Two commercially sourced powdered feedstocks were used in this work—one as thermal spray feedstock and the other as an erodent material for performance evaluation. An experimental powdered feedstock of Ti-6Al-4V from AP&C (Quebec, QC, Canada) was used for the HVAF deposition. The as-received powder particles were quoted to have a particle size distribution (D₅₀) in the range of 17–34 μm with a standard deviation of 10.9 µm. Alumina powder from AP&C (Quebec, QC, Canada) was used as the erodent for the erosion wear test performed on the as-sprayed coatings. The alumina powder was oven-dried at 150 °C for 24 h to remove inherent moisture. The substrate used for the coating deposition was Ti-6Al-4V, otherwise referred to as bulk Ti64 in subsequent sections. The substrates were grit-blasted with standard alumina particles to create mechanical interlock points on the substrates’ top surface.

2.2. Alloy Coating Deposition

The alloy coatings deposited from the Ti64 feedstock were completed using a deLaval convergent–divergent nozzle in an HVAF spray process setup (M3, UniqueCoat Technologies LLC, Oilville, VA, USA) running on propane fuel. The spray process utilized four different nozzle configurations—each configuration for a spray run (see

Table 1. Spray session and nozzle configuration.

S/N	Spray Sessions	Configurations
		Nozzle Length	Exit Size
1.	Run-1	Long	Wide
2.	Run-2	Long	Small
3.	Run-3	Short	Wide
4.	Run-4	Short	Small

)—to make four different coating samples. The spray parameters used during the spray session are presented in

Table 2. HVAF spray parameters for the four spray runs are marked by different nozzle types.

Spray Parameters	Spray Runs
	1	2	3	4
Nozzle type	4L4C	4L2C	3L4C	3L2C
Air pressure (MPa)	0.7	0.67	0.7	0.65
Fuel 1 pressure—propane (MPa)	0.65	0.67	0.65	0.68
Fuel 2 pressure—propane (MPa)	0.65	0.67	0.65	0.68
Carrier gas pressure—N2 (l/min)	40	40	40	40
Feed rate (g/min)	100	100	100	100
Pass velocity (m/min)	100	100	100	100
Pass spacing (mm/rev.)	5	5	5	5
Spray distance (mm)	200	200	200	200
Number of passes	10	10	10	10

.

2.3. Characterization and Performance Tests

2.3.1. Sample Preparation and Microstructure Analysis

The morphology and particle size of the as-supplied powder feedstock were examined in the secondary electron (SE) mode through the scanning electron microscopy technique using ZEISS Gemini field emission SEM equipped with an Oxford AZtec energy-dispersive (EDS) detector. A sparse distribution of the powder on the stub was gold-coated before the scanning electron imaging. The ZEISS SEM was also used to obtain micrographs of the coating top surface as well as the wear tracks/spots of the tested coating surfaces. The top surface micrographs, in SE mode, provide insight into the coating build-up, while the wear track micrographs provide insight into wear mechanisms. The micrographs of the coating cross-sections were obtained in backscattered electron (BSE) mode using a tabletop SEM (TM3000, HITACHI, Tokyo, Japan). The cross-section micrographs provide insight into the coating thickness and the possible defect (pores, cracks, and voids) distribution across the coating spread. The coating cross-section was prepared from hot-mounted, as-sprayed coupons sectioned using a precision cutter before the cut samples were polished according to standard metallographic procedures. The coating porosity was estimated from multiple cross-sectional micrographs using Image-J software (Image J 1.53e, Fiji, Dresden, Germany). The estimated porosity of the respective coating samples is reported with the associated standard error.

2.3.2. Microhardness

A microhardness test was carried out on samples of the coatings deposited from each spray run. The hardness test was conducted on a polished cross-section of the coating samples using a Vickers indenter (Duramin-40, Struers, Willich, Germany) at an indenter load of 200 gf sustained for 10 s. Each measurement consists of ten indents to report the average microhardness with the associated standard error.

2.3.3. Wear Performance Tests

The sliding wear test was conducted using the ball-on-disk configuration on the tribometer machine (TRB3, Anton Paar, Breda, The Netherlands) as per the ASTM G99 standard. The counter body was a stainless-steel ball of 6 mm diameter. The choice of stainless-steel counter body was to provide for in-service conditions. The load applied (P) onto the disk was 5 N to complete a total sliding distance of 100 m—see the schematic of the test process as shown in Figure 2. The specific wear rate (SWR) was obtained from the combined measurement of the wear track area and the track length using Profilm 3D white light interferometry (Filmetrics Europe GmbH, Unterhaching, Germany). The specific wear rate was thus the ratio of the volumetric material loss to the product of the total sliding distance and the normal contact load.

The erosion test was conducted on the Air Jet Erosion Tester TR-470 using alumina erodent powder. The erodent average flow rate was 2 g/min, and the impact velocity was set at 70 m/s for three different impingement angles (30°, 60°, and 90°)—see Figure 3. The choice of the impingement angle was to test the response of the deposited coatings under the envisaged in-service condition. Each erosion test was run for 2 min before the erosion rate for each impingement angle was recorded through a weight loss approach. The weight loss was measured using the machine MyCal from PCE instruments, model AB-100, with a sensitivity of 0.0001 g, while the surface roughness was measured using Surftest 301 from Mitutoyo (Kawasaki, Japan).

The erosion rate was calculated as the volume loss divided by the mass of erodent particles. Volume loss is the weight loss divided by the density of the coating, taken as 4.5 g/cm³. The mass of the erodent particles used per test session was estimated as the flow rate of the erodent multiplied by the test duration. In this case, 2 g/min multiplied by 2 min gives 4 g of erodent particles.

3. Results and Discussion

3.1. Microstructure: Materials and as-Sprayed Coatings

The secondary electron micrograph of the Ti64 powder shown in Figure 4 presents a loose powder with spherical morphology. The spherical morphology is indicative of the free-flow feedstock required for coating deposition. The particle size distribution plot is in Figure 4b, suggesting most of the particles are around 30 μm in size.

The erodent alumina material is presented in Figure 5. The secondary electron micrograph shows the particles are angular, with the bulk of the particles being in the size range of 40–70 μm, as shown in Figure 5b. The powder morphology suggests a reduced flowability, a reason for the prolonged oven drying of 24 h before the erosion wear test as given in Section 2.1.

Backscattered electron micrographs of the as-sprayed coatings are shown in Figure 6 and Figure 7. The images present the cross-section of the coatings, showing the coating substrate interface, inter-splat boundary, pores, and coating thickness. The microstructure of the four coatings shows a stack of unmolten individual particles with varied plastically deformed shapes. This observation suggests the as-received powder particles have lost their respective spherical morphology, as presented in Figure 4. Each of the coatings presents with a different porosity level via a visual inspection. A quantitative measure of the porosity for the individual coatings using image-J software suggests the coating from run-4 (see

Table 3. List of estimated porosity and measured thickness of the respective coatings.

Coating Features	Spray Runs
	1	2	3	4
Porosity (%)	2–3	0.5–1	1–2	<0.5
Thickness (mm)	0.65	0.32	0.32	0.30

) has the least porosity—less than 0.5%. Run-1 shows a porosity level of 2%–3%, which is the highest of the four coatings, while run-2 and run-3 have 0.5%–1% and 1%–2% porosity levels, respectively. Coatings deposited via HVAF have a high density resulting from the hypersonic impact velocity [2]. This feature of the HVAF spray technique engenders a compact overlay of plastically deformed, softened feedstock. The identified porosity on the cross-section of the coatings is formed from the inter-boundary particle geometries, however deformed; as such, the pores have irregular geometries depending on the morphology of the adjoining particles.

The characteristic features of a coating microstructure are also affected by the spray system hardware configuration [22]. The configuration of the nozzle used for each of the coating samples affects the microstructure of the coatings. The long nozzle with a wide exit meant longer dwell time for the particles in-flight, creating softened and/or molten particles that are deposited as splats. A reduced dwell time owing to the short nozzle length would mean the coating build-up elements will have more softened particles than splats; the reverse is also valid. The exit size affects the deposition velocity, and an enlarged exit means the coating build-up will rely on the thermal heating of the impacting particles rather than the impact energy due to particle velocity. The coatings sample in run-1 is built mostly based on the thermal energy of the particles because they have a longer dwell time inflight arriving at the substrate surface with reduced velocity through a wide exit, in contrast to the run-2 sample, which combined a longer dwell time due to the long nozzle length with a small exit. The former shows the highest porosity compared to the latter, which has the least porosity. This observation suggests the dwell time makes for the deposition efficiency of sprayed particles; it requires high impact velocity to build dense coatings with reduced porosity. The small exit configuration for the sample from run-4 further justifies the influence of the deposition velocity of the particles. The four coatings are deposited in the parameter window that produces mostly softened particles, as proved in the work of Browning [2]. It is a trade-off between deposition efficiency and coating density vis à vis the nozzle length and exit size; deposition efficiency drops with high impact velocity due to the bounce-off effect given that the feedstock particles are only softened and not molten. This explains the higher thickness of the coating sample in run-1 compared to the 50% loss of material in run-2 for the same number of passes and longer dwell time engendered by the long nozzle length. The deposition efficiency as a measure of the coating thickness is preserved in addition to the coating density with a short-length nozzle and a small-nozzle exit. This will have a significant productivity impact because a short nozzle length would mean coating deposition in a shorter time.

The top surface morphology of coatings reveals the makeup of a thermal spray coating. It shows the spread of deposited materials when molten [23,24], clusters of materials that may be partly molten, or materials retained from the sprayed feedstock [25]. The four coating samples (runs 1–4) show consistency in this regard; the top surface of each of the coatings presents mostly highly deformed materials. A representative micrograph of the coating’s top surface is shown in Figure 8. The coating buildup consists of layers of deformed materials well distributed to cover the entire substrate’s top surface. The key features of the deformed material overlays are shown in the inset, which reveals the inter-particle boundary, microcrack, and blind pores. A magnified void in the top right-hand corner of Figure 8 shows a buried, undeformed particle from the sprayed feedstock. The inter-particle boundaries of the deformed materials are close, suggesting good interlayer bonding. There are, however, some micro-cracks; those would have developed from cooling transformation stresses.

3.2. Microhardness and Wear Performance

3.2.1. Microhardness

Given the varied microstructural build-up of the individual coatings as shown by the respective cross-sectional views presented in Section 3.1, it is expected that each of the coatings will show different resistance to local plastic deformation measured using Vickers hardness. Figure 9 shows the measured hardness of the respective coatings with the associated standard error. However, the upper and lower limits for all the samples lie in the range of 320 and 400 HV. The processing condition seems not to cause any significant variation in the microhardness of the samples. The dominant microstructural defect in all the coatings is porosity, with varied percentages as presented in Table 3. Nonetheless, all the coatings show similar resistance to local plastic deformation. In each of the coatings, the indentation energy was dissipated in one or both ways of absorption through particle spreading to fill inter-boundary pores or dissipation through inter-boundary cracks for adjoining particles/splats.

3.2.2. Sliding Wear

The specific wear rates of the coatings relative to the bulk Ti64 sample are shown in Figure 10. Each of the samples shows a wear rate of the order of 10⁻⁴ mm³/Nm—the respective magnitude of wear shown by the individual samples lies in the same standard error range. However, the wear regime and the dominant wear mechanism for each sample will be different. Wear regimes identified for dry sliding wear conditions are either mild or severe; each one has its own dominant features. In mild wear, mixed oxide layers form and delaminate, along with the delamination of the layers next to the contact surfaces. On the other hand, severe wear causes plastic deformation due to the thermal softening of one of the surfaces in contact [26]. The features shown by the wear regimes define the mechanism that drives the wear of the surfaces in question. The sliding wear test presented here has a stainless-steel counter body against a metal alloy; the metallographic make-up of the two surfaces in contact suggests a hard-against-a-soft body interaction. Given this condition, it is expected that the tested samples will experience mild or severe wear, depending on the dominant wear mechanism.

The wear tracks of the samples shown in Figure 11 present features that define the dominant wear regime and mechanism for each of the tested samples. The features shown for each of the tested coating samples reflect the measured properties of the respective coatings as well as the microstructure. Run-1 has the highest specific wear rate as estimated, and its worn surface shows deformed layers, deep plowing grooves, fatigue cracks, and piles of transferred materials. The dominant wear mechanism seems to be plowing given the deep groves that dominate the worn track. Where the materials are not removed, they have undergone plastic deformation; the repeated contacts between the surfaces also subjected the layers to fatigue loading, causing cracks. All the coatings show identical wear mechanisms, but one form is dominant in one compared to others. The height of the plowing grooves reduces from run-1 to run-4; deformed layers dominate the worn track for run-2, with the most distribution of fatigue cracks amongst the samples. The worn rack of run-4 appears to have the shallowest grooves when compared via visual inspection to other worn tracks. The run-4 wear track, however, has large masses of transferred materials across its length, with some randomly distributed fatigue cracks. In any of the cases, none of the coatings showed severe material removal to expose the underlying substrate. Where materials were moved, the transferred materials adhered to the wear track to ensure overall material conservation. This observation suggests the coatings would support the lifespan elongation of a work part coated for repair.

3.2.3. Erosion Wear

The erosion wear rate of the four coating samples compared to the bulk Ti64 sample is presented in Figure 12. The erodent material has been impinged at the surfaces at three different angles to show the effect of the impingement angle on the erosion rate. In-service conditions of engineering components may expose them to randomly impinging particles. This is particularly true for aeroengine parts operated in sandy environments [27,28]. The contact angle between the impinging erodent particles and the surface of the engineering component is an extrinsic factor to determine the resistance of the impacted surface [29]. Angular particles hitting a surface at an acute angle deliver contact energy sufficient to cause material deformation due to ductility, indents, micro-cracks, or material removal through micro-chipping or plowing, depending on whether the material is ductile or brittle [30]. All the coatings as well as the bulk Ti64 sample show an increased wear rate as the impingement angle goes acute from 90° to 30°. The highest erosion rate is present at the 30° impingement angle for all the tested samples; this is consistent with the observation of previous work on Ti-6Al-4V [29]. However, the least porous of the coating samples, run-4, was expected to mimic the bulk Ti64; it shows superior erosion wear resistance compared to the bulk sample.

The erosion wear response of materials scales with the metallographic makeup of the material, depending on whether the material is ductile or brittle. Ductile materials show a higher erosion wear rate at perpendicular erodent impacts compared to acute angles; brittle materials show the direct opposite of this behaviour [23]. Ductile erosion behaviour is characterized by micro-plowing, cutting, and/or scratching [30]. The mechanisms of wear known with Ti-6Al-4V include plastic deformation, micro-plowing, and micro-scratching [29]. The representative overviews of the impact sites of the erodent particles on the surface of the coatings are shown in Figure 13. The elongated oval geometry created via erodent impacting at 30° suggests a unidirectional flow of material in congruence with the direction of the applied force. The perpendicular impact of the erodent particles hitting at 90° created a circular geometry, while the erodent particles hitting at 60° created a bulged circle.

The magnified view of each of the erodent sites was not significantly different for each of the tested samples, coatings, and bulk Ti64 alike. This is despite the observed microstructural build of the respective coatings. The dominant mechanism However, it varies with the angle of impact for the erodent particles, which must have created the characteristic overview for each of the erosion test conditions. Representative micrographs for the three tested impingement angles are shown in Figure 14. The identified mechanisms are marked in addition to a schematic representation for each of the mechanisms. The influence of the two acute angles of impingement (30° and 60°) is shown as the micro-plowing tracks marked “3” and the micro-scratches marked “2”. These two mechanisms dominating the sites as presented corroborate the erosion wear rate results. As a ductile material would show micro-plowing and micro-scratching, such combined features would culminate in an overview of the bulged circle and an elongated oval shape. The perpendicular impingements left diamond-like imprints on the coating surface—see the points marked “4”, Figure 14d. All the mechanisms elicited by the coatings show the retention of the coating material on the substrate, irrespective of the repeated shear stress created at the impact points of the erodent particles. Ti-6Al-4V bulk material is known to show a maximum erosive material removal rate at acute angles [29]. The one closest to a bulk material in terms of microstructure among the four coating samples is run-4. The run-4 coating interestingly shows an improved erosion wear rate compared to the bulk sample. This gives a useful indication of the potential of a dense coating to function well as a repair layer for worn parts. Given the deposition configuration of the run-4 sample, the short nozzle length and short nozzle exit mean the particles have reduced in-flight resident time while exiting at higher temperatures than particles deposited from a wide exit nozzle. This implies the particles only became heated enough to ensure high deformation for a high packing density, as confirmed by their porosity of <5%. This sample therefore has its predominant microstructural features as inter-particle boundaries instead of pores. Grain boundary strengthening is a known mechanism to improve the strength of materials. When the inter-particle boundaries are combined with the strain rate that could have accompanied the particle deposition from the hypersonic speed of the HVAF process, the run-4 sample may have improved strengthening due to grain boundary and work hardening because of the accompanied strain rate of its buildup particles [31]. These strengthening mechanisms may explain the improved erosion resistance of the run-4 sample over the bulk Ti64 sample.

4. Conclusions

This paper presents the HVAF deposition of coatings using commercially sourced Ti-6Al-4V feedstock powder sprayed with different nozzle configurations. The microstructure, sliding wear, and erosion wear of the coatings have been investigated to reach the following conclusions:

• The HVAF technique is capable of depositing Ti-6Al-4V coatings with very high density using an appropriate nozzle configuration;
• The microhardness of the coatings is similar, which correlates with their sliding wear performance;
• The densest of the coatings (<0.5%) compared well with the bulk Ti-6Al-4V material under both sliding and erosive wear conditions and showed improved erosive performance over the bulk Ti-6Al-4V sample;
• The measured properties and the wear performance of the tested coatings are suggestive of the promise of HVAF deposition of Ti-6Al-4V coatings for component repair.