The Effect of HVOF Spray Distance on Solid Particle Erosion Resistance of WC-Based Cermets Bonded by Co, Co-Cr and Ni Deposited on Mg-Alloy Substrate

Magnesium alloys are very interesting engineering materials because of their high strength-to-density ratio. On the other hand, they are characterized by low hardness as well as low erosion resistance. Because of these reasons, their applications in the industry are very limited. The article presents the results of the high velocity oxy-fuel (HVOF) spraying of the hard cermet coatings onto AZ31 magnesium alloy substrate. Three feedstock powders were used in the process with composition (wt.%): WC-12Co, WC-10Co-4Cr and WC-20Cr 3 C 2 Ni. The spray distance (SD) was selected as a variable parameter with values equal to 320 and 400 mm. Observations carried out under a scanning electron microscope (SEM) revealed a typical HVOF-sprayed microstructure with a compact structure and low po - rosity (below 3 vol.%). The hardness of the manufactured coatings, ranging from 912 HV0.2 to 1328 HV0.2, what was significantly higher than the substrate. The solid particle erosion tests were carried out according to the ASTM G76-04 standard. Erosive experiments were done for 30°, 60° and 90° inclination angles of the nozzle using Al 2 O 3 abrasive. Erosion tests confirm that cermets exhibit substantial erosion resistance better than the substrate. The highest erosion resistance was noted for WC-10Co-4Cr coatings. The erosion rate for cermet coatings was mostly below 0.9 mg/min, whereas for the AZ31 it was more than 1.5 mg/min. In the case of the average erosion value, it was between 12 and 22 times lower than for the substrate. Results analysis reveal that shorter spray distance decreases porosity, increases hardness, and finally supports erosion resistance of the cermets.


INTRODUCTION
Current expectations in the industry are related to constantly growing requirements regarding device parameters, extended operating time and cheaper service.Such multi-dimensional expectations are not easily achievable, especially in a situation of shrinking natural resources and rising prices at the same time.Therefore, we can increasingly observe situations in which, instead of focusing on inventing new materials, scientists and engineers try to improve existing ones.Research on magnesium alloys fits into these trends [1][2][3].The most important advantages of such material are low density and, consequently low value of the specific strength [4].On the other hand, there are crucial drawbacks which defects that prevent the wide use of these materials, e.g.low hardness, poor corrosion-, wear-and erosion resistance [5,6].
In order to improve the surface of the magnesium alloys, the two routes are possible.The first option is surface modification.It could be done by many methods, however part of them is not suitable for magnesium alloys.The valuable review of this area was given by Morelli et al. [7].The second way is connected with coatings deposition using thermal spray methods.In such case there are some limitations connected with the flammability of the magnesium alloys [8] and high plasticity [9] from the one hand.On the other hand, there are some limitations according to the methods and feedstock materials, e.g., in case of the cermet coatings deposition.Among many thermal spray methods, the high velocity oxy-fuel (HVOF) is characterized by homogenous structure with low porosity level, as well as low crack density [10][11][12].It is caused by relatively low flame temperature and high particles velocity [13][14][15] in comparison to atmospheric plasma (APS) and suspension plasma spraying (SPS) [16,17].
In the available literature on the discussed issue, the aspect related to the difficulty in selecting the parameters of the process of spraying cermet coatings on a relatively soft substrate material, especially aluminum alloys is repeatedly described [18][19][20][21].Usually structural stainless steel or nickel alloys are commonly used as substrate materials [12,22,23].These alloys show relatively high melting points, which facilitates the selection of HVOF process parameters to obtain high-quality coating and prevent melting down the substrate.In the case of aluminum or magnesium alloy substrate, the deposition of high-quality cermet coatings seems a tricky task that must solve the problem to achieving sufficient spraying temperature (to liquidate the cermet feedstock cobalt or nickel-based matrix) and maintaining the temperature of the substrate below its melting point.This makes the appropriate selection of deposition parameters on magnesium substrate challenging, especially in the case of Co-WC and Ni-WC based cermets manufactured on AZ31 magnesium alloy.Successful selection of the technological spraying parameters was shown in the current research.
HVOF coatings deposited on magnesium alloy substrates could be a promising solution in the automotive and aerospace industries because they allow for mass reduction.Some works in the literature were made in the field of corrosion resistance of cermet coatings [22,24].Nevertheless, a literature gap exists in cermet coatings manufactured on magnesium alloys and their deep investigations to verify the operational properties of cermets deposited using specific technological parameters.One such way is a solid particle erosion resistance.The second issue is the selection of the key HVOF process parameters and their impact on microstructure and such properties.One of the most important parameters is spray distance, especially in industrial conditions, which is relatively easy to change [25].As mentioned before, the appropriate selection of spray distance guarantees the high quality of the deposited coatings and prevents the magnesium substrate from melting, which was revealed in the current study.This research results are part of a broad experimental plan regarding the properties of coatings thermally sprayed using the HVOF method on light material substrates such as magnesium alloys of the AZ31 series.The work is a continuation of research aimed at determining the impact of the applied process parameters on the functional properties of the applied coatings.In our previous work, corrosion resistance, abrasive wear and cavitation erosion were tested [26,27].Moreover, the detailed microstructure investigations were carried out [28].Testing the resistance of coatings to erosive wear caused by solid particles is the next stage of the research plan.
Current investigations aimed to assess the influence of microstructure, hardness, and surface morphology on solid particle erosion resistance of cermet coatings: WC-12Co, WC-10Co-4Cr and WC-20Cr 3 C 2 -7Ni manufactured by HVOF on AZ31 magnesium alloy substrate using different spraying distance equal to 320 mm and 400 mm.Additionally, the mechanisms of erosion wear of deposited coatings were identified.This paper fills the literature gap in state of the art on cermet coatings deposited on the light materials substrate.Also it provides the original comparison of erosive resistance of a set of different nickel and cobalt-based binder cement systems sprayed on AZ31 substrate.

MATERIALS AND METHODS
In current studies three commercially available cermet powders: (i) WC-12wt.%Co(Amperit 558.074,Höganäs), (ii) WC-10wt.%Co-4wt.%Cr(Amperit 518.074,Höganäs) and (iii) WC-20wt.%Cr 3 C 2 -7wt.%Ni(Woka 3702-1, Oerlikon Metco) were used as feedstock in the deposition process.For all powders the delivery conditions were the same, agglomerated and sintered.The feedstock powders morphologies are presented on Figure 1.For all powder types the delivery state was agglomerated and sintered.Moreover, the particles exhibit spherical shape which is preferential from the its flowability point of view.
The particle size distribution declared by the manufacturer (-45 + 15 µm) was confirmed by own measurements (see Table 1) carried out with particle size analyzer set-up (PSA 1190, Anton Paar).The detailed analysis of the feedstock powders can be found in [26].
The substrate material used in current study was magnesium alloy AZ31.The specimens were prepared as a circle with 100 mm diameter and 5 mm thickness.Before spraying the substrates were sand blasted in order to obtain specific surface roughness (Ra about 3 μm) and cleaned with ethanol.The deposition process was carried out using a spray system JP 5000 TAFA (Indianapolis, USA) from RESURS Company (Warsaw, Poland).The constant process parameters were equal to as follow, oxygen feed rate: 900 (slpm), kerosene feed rate: 0.435 (slpm), nitrogen feed rate: 12 (slpm), powder feed rate: 70 (g/min) and water flow: 23 (slpm).The spray distance (SD) was selected as a variable parameter.Sample codes, as well as the SD value variations, are collected in Table 2.
The microstructure analysis of the manufactured coatings was carried out Scanning Electron Microscope with secondary electron and backscattered detectors (Supra 35, Zeiss, Oberkochen, Germany) which was equipped with EDS (Energy Dispersive X-ray Spectrometer) analyzer (Supra 35, Zeiss, Oberkochen, Germany) in order to determine chemical composition.Before the tests, additional calibration of the EDS system was performed on real standards for all elements, including carbon.Image analysis with free software ImageJ (version 1-50) was used to determine the porosity level.The measurements were made according to the ASTM E2109-01 standard.The porosity was assessed as an average from 20 images at 1000× magnification.Scanning Electron Microscope with secondary electron and backscattered detectors (Supra 35, Zeiss, Oberkochen, Germany) was used for topography investigations of the deposited coatings.The surface roughness of sprayed coatings was measured by a stylus profilometer (MarSurf PS 10, Mahr, Germany), according to the ISO 4288 standard, with Gaussian filters according to the ISO 16610-21 standard.Five Ra and Rz parameter measurements were performed for each sample.Then, calculate the average and standard deviation values.The Vickers hardness was estimated at the sample cross-sections under the load equal to 1.96 N (HV0.2) using the HV-1000 hardness tester (Sinowon Innovation Metrology, Dongguan, China), according to the ISO 4516 standard.For each coating, 12 indentations were made and then the average and standard deviation values were calculated.
Erosion resistance tests at room temperature were carried out on a specially designed test stand at the Welding Department of the Silesian University of Technology in Gliwice according to ASTM G76-04 standard.Figure 2 shows the schematic diagram of the erosion resistance test rig according to the ASTM G76-04 standard.
The test was performed using a nozzle tube with a diameter of 1.5 ± 0.075 mm and located 10 ± 1 mm from the sample.The inclination angles of the nozzle relative to the sample were 30°, 60° and 90°, respectively.The test uses Al 2 O 3 abrasive with a particle size of 50 μm (Figure 3), the particle feed amount was 2 ± 0.5 g/min, and the nozzle exit speed was 70 m/s.The test duration was 10 minutes, all tests carried out in the assprayed state.To obtain reliable results, 10 tests were performed using the given parameters.The quantitative result of these tests was the mass loss using a laboratory scale with 0.1 mg accuracy.According to the ASTMG76-04 standard, the erosion rate and the erosion value were calculated.Moreover, the SEM observations were carried out in order to determine the wear mechanism.

RESULTS AND DISCUSSION
All coatings were deposited in order to achieve a thickness c.a. 250 µm.As expected, the coatings exhibit a dense and homogenous structure.Moreover, the good filling of the substrate surface irregularities could be seen (Fig. 4).It is a result of the specific conditions, especially high velocity of the molten powder particles, which is characteristic for the HVOF method.Similar structure could be found in [29,30].
The detailed observations at higher magnification (Fig. 5) confirmed the fine structure of the all coatings.The tungsten and chromium carbides are uniformly distributed in the hole samples volume.On the other hand, both metallic matrixes were melted and correct wetted hard particles.Some imperfections could be observed, as cracks, pores and voids.It is probably a result  of the thermal stresses caused by the deposition process [31] as well as short time in which the particles stay in the relative low-temperature flame (c.a.3200 K) [32,33].The analysis of chemical composition were carried out on the polished cross-sections of tested coatings, in the points marked in Figure 6, in order to confirm the presence of elements.The EDS results are collected in Table 3 and were calculated using the ZAF correction.
The porosity values of the manufactured coatings were at the level for typical HVOF deposits.Even for samples sprayed from 400 mm, the porosity was relatively low (below 3 vol.%).Detailed results are collected in Table 4. Similar results could be found in [34][35][36].Two reasons may cause the differences in the porosity values for investigated coatings.First is a type of matrix material.Nickel is a slightly softer metal than cobalt and has a lower melting point [37].On the other hand, longer spray distance (SD) makes it impossible to obtain a compact structure because of reduced particles velocity [38].In Table 4, the influence of the spray distance on the coatings' porosity value is clearly visible, which was also reported by Murugan et al. [39].
The hardness of the manufactured coatings is presented in Figure 7.It could be seen that reducing the spray distance plays an important role in improving hardness.It is a result of more compact structure as well as reduced porosity [40,41].Similar hardness values to those in current studies could be found in [42,43].The surface of the obtained coatings is rather smooth for standoff distance (SD) equal to 320 mm (Fig. 8

left column).
At some areas, minor irregularities could be observed.In case of SD equal to 400 mm (Fig. 8 right column) there are some craters which caused increasing surface roughness.Nevertheless, for all manufactured coatings their topography is typical  3. EDS results of the investigated coatings (spots marked in Fig. 6), SEM   for HVOF deposits [44,45].The observed irregularities are connected with carbide particles that exhibit high melting points and remain solid during flight in the flame [46].The surface roughness of the as-sprayed coatings is presented in Table 5.These values are relatively close to the literature data [34,47].The coatings with cobalt matrix show a slight decrease in roughness with the increasing SD.It could be caused by decreasing particles velocity and, consequently, kinetic energy [48].It results in smaller surface deformations.
In general, the surface roughness values (for both factors) differ slightly in the standard deviation range.Therefore, there was no clear relationship between the spray parameters, feedstock material type and the surface morphology of coatings.
The results of the solid particle erosion tests are presented in Figure 9.It is clearly observed that HVOF coatings strongly improved the erosion resistance in comparison to the substrate material.As can be seen, the average erosion rate (AER) for all coatings is around double, sometimes even 2.5 times, lower than for AZ31.Moreover, the results confirmed that for bulk metal alloys, the most dangerous impact angle is 30°, whereas for ceramic materials, it is 90° [49].In case of the cermet materials, the maximum erosion rate is connected with the impact angle c.a. 60° [50].Analysing Figure 9, it can be seen that for all impact angles, as well as all coating materials, the lower SD results in better erosion resistance.Among tested samples, the best performance exhibits C2-320.The excellent erosion resistance is characteristic in the case of WC-10Co-4Cr coatings [51,52].The novelty of current investigation is confirmation assumed hypothesis that changing substrate material onto the light Mg-based alloy with correct process parameters makes possible obtaining coatings without deterioration of its properties.In Figure 10 were presented the results of the average erosion value (AEV).It is calculated as material volume loss in relation to the total mass of erodent.In this case, the impact angle equal to 60° was also the most hazardous for the cermet coatings.Moreover, the differences between coatings and AZ31 substrate are much and more significant than for AER due to almost seven times higher density of cermet materials than the magnesium alloy.For better readability of AEV values for cermet coatings, its results were presented in separate plot (Fig. 11).
Erosive results and surface morphology and analysis indicate a negligible effect of surface roughness (Tab.5) of as-sprayed coatings on the erosion rate (Fig. 9).On the other hand, research confirms that the hardness and uniformity of the coatings have a crucial effect on the erosion resistance of the coatings.It can be deduced from Table 4 and Figure 7 that a shorter spray distance i.e. 320 mm, decreases the porosity and increases    ach coating's hardness.Combining the hardness (Fig. 7) with erosion results (Fig. 9; Fig. 10) confirms that coatings sprayed from 320 mm standoff distance are harder and show a lower material loss rate than those deposited with 400 mm distance.The exemplary eroded craters, due to solid erosion particle tests at different impact angels, were compared in Figure 12 (for C3-320) and Figure 12 (for C2-320).At an angle equal to 30°, the main erosion process is associated with microcutting and grooving (lip formation) of the coating material (Fig. 12b; Fig. 13b).During the test, the carbide particles are exposed because of the matrix removal.It results in the displacement and, finally detachment of these particles, especially under the small impact angle.Similar behaviour was observed by Gonzalez et al. [50].In case of impact angles equal to 60° and 90°, the erosion mechanism was changed.The main reason for the material loss is the formation of the craters.This behaviour is characteristic for brittle materials [53].Nevertheless, the most eroded surface, with many pits, empty areas after carbide removal due to fatigue action of erodent, microcutting and cracks are visible, which is a characteristic feature for an impact angle equal to 60° (Fig. 12c,d and Fig. 13c,d).Lip formation and brittle behaviour result from the mixed nature of cermet, consisting of metallic binder and WC ceramics [54].
To summarize, cermet coatings' solid particle erosion mechanism relies on microcutting and grooving the metallic matrix, lip formation, detachment of carbides, and brittle cracking [50,55].Compared to the samples deposited using a spray distance 320 mm, much severe deterioration has been observed for coatings sprayed from larger standoff (400 mm), resulting in lower hardness and uniformity obtained for cermet's sprayed with 400 mm.The longer spray distance influences the less compact structure and the reduced cohesion in the coating.

CONCLUSIONS
The main research goal of the current paper was to investigate the influence of spray distance and type of feedstock material on the erosion resistance of the WC-12Co, WC-10Co-4Cr and WC-20Cr 3 C 2 -7Ni cermet coatings deposited on magnesium alloy substrate by HVOF method.The successful manufacturing of the high-quality cermet coatings on the light AZ31 substrate was proved.Analysis of the results shows that spray distance is one of the key process parameters and influences the coating density, hardness and erosion resistance.On the other hand, the feedstock type negligibly influences these properties.The following findings can be summarized: • the thickness of the deposited coatings was c.a.
250 µm and all the coatings have a dense and uniform structure.In addition, the irregularities of the substrate surface were well filled.Some defects such as cracks, pores and voids can be observed as a result of the HVOF thermal spray process; • the porosity values of the manufactured coatings are at the level of typical HVOF deposits.Even for samples sprayed from 400 mm distance, the porosity is relatively low (up to 3 vol.%).Generally, coatings sprayed with a distance of 320 mm show lower porosity and higher hardness.C3-400 coating has the lowest hardness (912 HV0.2) and C1-320 has the highest hardness (1328 HV0.2).Reducing the spraying distance plays an important role in improving hardness.This is due to a denser structure and, consequently, lower porosity; • the surface of the manufactured coating is relatively smooth for the coatings deposited, with a standoff distance equal to 320 mm.Minor irregularities can be observed in some places.At the spraying distance equal to 400 mm, some micro craters appeared, resulting in an increase in surface roughness.C3-400 coating has the highest roughness (Ra = 4.60 ± 0.34 µm).The roughness decreases slightly with increasing spraying distance for coatings containing cobalt matrix.There was no clear relationship between the spray parameters, surface morphology of as-sprayed coatings, and erosion resistance; • it can be clearly seen that the HVOF coating significantly improves the erosion resistance compared to the substrate material.The lowest average erosion rate (AER) is characterized by the C2-320 coating (0.54 mg/mm) with an impact angle of erodent equal to 30°, while the highest is the C3-400 coating (1.04 mg/mm) with an impact angle equal to 60°.Also, for average erosion value (AEV), the highest (approximately 0.0407 mm3/g) is characterized by the C3-400 coating with an impact angle of 60° and is the most hazardous for all cermet coatings.Dominant erosion mechanisms of cermet coatings were stated.Erosion relies on microcutting and grooving of the metallic matrix, material fatigue leading to brittle cracking, carbides and coating material removal; • increasing each cermet hardness improves the erosion resistance.This depends on the microstructure and HVOF spray distance.A shorter spray distance i.e. 320 mm decreases each coating's porosity and increases hardness, which both facilitate the erosion resistance of the coatings.In other words, all cermet coatings sprayed from 400 mm standoff distance show higher material loss than those deposited with 320 mm distance.

Figure 7 .Figure 8 .
Figure 7.The average hardness values of the deposited coatings

Figure 9 .
Figure 9.The average erosion rate of the sprayed cermet coatings and the AZ31 substrate

Figure 10 .Figure 11 .Figure 12 .
Figure 10.The average erosion value of the sprayed cermet coatings and the AZ31 substrate

Table 2 .
Sample code and values of the variable HVOF process parameters

Table 4 .
Average values of porosity for the manufactured coatings (in vol.%)

Table 5 .
The average roughness value (Ra and Rz) of the manufactured coatings