2016 Volume 63 Issue 7 Pages 688-696
In the present work, three experimental WC-17 Co plasma sprayed coatings were developed from submicrostructured, conventional and bimodal (composing of 30 wt.% submicrostuctured and 70 wt.% conventional WC grain) agglomerated powder. The effect of carbide size distribution on the microstructure, phase structure and properties such as microhardness, fracture toughness and sliding wear performance of three coatings were measured. The experimental results indicated that bimodal coating had little porosity of microstructure with lower degree of decarburization of the WC phase to W2C and Co6W6C phases, consequently, showed both high microhardness and fracture toughness, e.g. the microhardness is similar to that of the submicrostuctured one and the fracture toughness is comparable to that of the conventional one. The coefficient of friction for the coatings decreased slightly with increasing test load, and that friction coefficient of bimodal coating was lower than the other ones. The wear loss of both submicrostuctured and conventional coatings were higher than the bimodal coating and increased with increasing test load. The microscopic analyses of the wear surface showed the cobalt extrusion followed by carbide fracture, or pulled out were the predominant failure mechanism. Besides that, fatigue cracking of the submicrostuctured coating at a high test load played an importance on the wear mechanism.
In the field of protecting surfaces against wear, WC-Co based coatings with higher sliding, abrasive and erosion wear resistance are used extensively for numerous wear resistance applications such as mid-span stiffeners (shrouds), pump seals, sucker rod couplings, extrusion dies, aircraft flap tracks, exhaust fans and crushing rollers, etc1–5). It is well known that the mechanical properties of WC-Co coatings are closely related to microstructure characteristics such as porosity, phases distribution, carbide size and splats cohesion6). Numerous studies have concluded that the hardness increases and fracture toughness decreases for WC-Co coating with decreasing carbide size6–8). Therefore, the coating with finer WC particles has better sliding wear resistance due to its higher hardness proposed by Morks8). However, it is argued by Shipway9) that nanostructured coating with higher microhardness has lower sliding wear resistance than the conventional coating, which the sliding wear resistance is related to fracture toughness. Therefore, it should be noted that suitable selection of WC size is a significant benefit for improving the fracture toughness and hardness of WC-Co coatings.
Several studies10,11) have shown that a nanostructured WC crystal results in higher hardness. However, there are some difficulties in the obtention of enhanced fracture toughness in these thermal sprayed coatings as carbide size decreases, due to the extensive decarburization of the nanostructured WC-Co powder experienced during spraying, diminishing the sliding wear resistance at a higher load. Bimodal powders with the mixing structure of finer and coarse carbide particels have been produced to minimize this effect, keeping enhanced properties when compared with conventional ones12). As the finer component is the minor constituent not only the decarburization could be reduced, but also the total cost. To date, the microstructure and carbide size distribution of HVOF sprayed multimodal WC-12 Co coating have been investigated in great detail. However, the detailed microstructure and carbide size distribution of atmospheric plasma sprayed WC-17 Co coating is little understood.
In the present work, submicrostructured, conventional and bimodal as-received powders are fabricated by agglomerated and sintered technology. The microstructure parameters (porosity, phase distribution and mean binder free path, etc) of atmospheric plasma sprayed WC-17 Co coating were characterized, with various carbide size. The microhardness, fracture toughness and sliding wear behavior of these deposits were experimentally studied. Finally, the effect of WC size distribution on the microstructure and properties of the deposits was discussed.
Three powders used to deposit the coatings in the present work, from BGRIMM Advanced Materials Science & Technology Co. Ltd, China, were submicrostructured (labeled S), bimodal (labeled B) and conventional (labeled C), respectively. All the Powder aggregates were manufactured by spray drying an aqueous slurry of WC and cobalt, and then sintered into sprayable particles in the range 15–45 μm for the atmospheric plasma spray (APS) process13). Bimodal powder was synthesized by mixing coarse 2.6–4 μm with fine 0.6–0.9 μm WC particles, wherein the ratio of coarse/fine was 70/30. However, monomodal WC particles used to form submicrostructured S and conventional C aggregates were fine and coarse powders, respectively.
Prior to spraying, carbon steel substrates (AISI 1045), degreased with acetone, were grit blasted using 20 meshs Al2O3 particels, with a mean roughness (Ra) of 6 μm. S, B and C grade powders were deposited onto two types of carbon steel substrates: flat substrates (60 × 30 × 2 mm) and cuboid substrates (19 × 12.35 × 12.35 mm), using an atmospheric plasma spray process (GTV-MF-200-APS-F6, Germany). To minimize variations in spray parameters, all the Powders were sprayed with identical conditions (presented in Table 1). The coating thickness was approximately 400 μm at the spraying condition, followed by grinding and polishing to 300–330 μm for microstructure observation and sliding wear test.
| Spray parameters | Values |
|---|---|
| Current [A] | 660 |
| Voltage [V] | 68 |
| Argon flow rate [L/min] | 60 |
| Hydrogen flow rate [L/min] | 5.1 |
| Powder feed rate [g/min] | 38~50 |
| Spray distance [mm] | 130 |
The characterizations of the as-received powders, coating cross sections and worn surfaces was studied utilizing SEM and backscattered (BSE) electron microscopy (JSM-7001F, Japan). Coatings porosity was also measured utilizing a optical microscope (Olympus BX6, Japan) coupled with an image analysis software (Image Pro plus 6.0, Media Cybernetics) on polished cross-sections of the coating at a magnification of X 200. Quoted values were an average of 20 areas for each coating.
Phase structures of as-received powders and as-sprayed coatings were studied using X-ray diffractometer (Ultima IV, Ragaku), at 40 kV and 40 mA, over 2θ = 20–90° with Cu Ka radiation and step 0.02°. The relative ratios of the WC, W2C and Co6W6C phases were calculated by RIR method14). Hence, the quantitative analysis did not include amorphous/nanocrystalline such as Co in the calculation.
The average diameter of WC grains (dWC) and the mean free path of the binder (LCo) were measured on the polished cross sections by quantitative metallography and stereology of linear analysis15,16). Equations (1) and (2) were used to calculated dWC and LCo for analysis.
| (1) |
| (2) |
where VWC is the volume fraction of WC phase, NWC/WC is is the number of noncontiguous boundaries between WC grains on a metallographic image by a line of unit length, NWC/Co is is the number of noncontiguous boundaries between WC grains and cobalt by a line of unit length. For each coating, 20 evenly distributed backscattered electron images were analyzed on the cross-section for each sample to determine VWC, NWC/WC and NWC/Co.
2.3 Mechanical Properties MeasurementsCross-sectional microhardness measurements were carried out on the transverse section of the coating by a Vickers microhardness tester (TUKONTM 5000, America), under a 2.94 N load for 15 s. Values quoted were an average of 30 indentations for each sample.
The fracture toughness was conducted based on an indentation method described in Refs17). A Vickers indenter was used on cross-sections of the coatings at 4 kg load with 15 s dwell time. Equations (3) was used to calculated fracture toughness (KIC), proposed by Evans & Wilshaw17).
| (3) |
Where P is the load, a the half-diagonal of the Vickers indentation and c is the length from center to crack tip. To increase measurement accuracy, no fewer than 10 indentations were made for each sample.
2.4 Sliding Wear TestThe dry sliding behavior of three coatings was invstigated according to Chinese National Standards (JB/T 9396-2013) using a block-on-ring tribological tester (MR-H5 tester, China)18) as shown in Fig. 1. Sliding testing was conducted on cuboid specimens coated, with parameters listed in Table 2, using GCr15 bearing steel (57–58 HRC) as counterpart. The relative humidity and the temperature were constant at <50 % and 25 °C, respectively. The coefficient of friction was plotted as a function of test time. In order to quantify the weight loss of the coatings during the block on ring test, all of the coated specimens were first ultrasonically cleaned in acetone, dried, and weighed to ±0.01 mg precision (sartorius cubis® MSA225S-0CE-DA, Germany) before testing, and then cleaned to determine the weight loss after testing.
Schematic illustration of the sliding wear test system of coatings.
| Sliding wear test condition | Values |
|---|---|
| Load [N] | 50, 150 |
| Sliding Speed [rpm, m/s] | 400, 1.03 |
| Sliding Duration [s] | 3000 |
Fig. 2 shows SEM three dimensional images of the as-received agglomerated powders. It can be seen that three powders exhibit near-spherical particle shape and uniform particle distribution. The differences between WC grain size of the agglomerated particles can be observed clearly in the three enlarged images. The submicrostructured powder with smooth surface, has a WC size of 0.6–0.9 μm as shown in Fig. 2b, whereas the bimodal and the conventional powders with rough surface, have wider WC size of 0.6–4 μm and 2.6–4 μm, respectively. A very porous structure was observed for the bimodal powder in Fig. 2c (global view), which was the result of wider WC size range. During thermal spraying, porous structure would contribute to high thermal input to heat the powder particles to semi-molten state, proposed by Ding19). The XRD patterns of the initial phases of the agglomerated powders revealed no differences. All of them were pure, containing only the required initial phases: WC and Co as shown in Fig. 3.
SEM images of the submicrostructured (2a, 2b), bimodal (2c, 2d) and conventional (2e, 2f) as-received powders.
XRD patterns of the the submicrostructured (S), bimodal (B) and conventional (C) as-received powders.
The XRD patterns of the coatings with different carbide size are presented in Fig. 4. It is evident from three patterns that there is on difference between phase compositon of all coatings, with the exception of varying peak intensiy. Major WC and minor W2C, Co6W6C were observed in the three patterns, similar to that reported in the literature20). This can be attributed to decomposing/decarburization of the carbide particles during the plasma spray process. It is seen that an amorphous hump at 2θ angles of 33–45° observed in the patterns, as has been shown in the Guilemany study for HVOF sprayed WC-12Co coating21). It suggests the molten droplets are estimated to cool at a rate in excess of 106 K.s−1 during the plasma spraying, which is suitable for forming an amorphous phase.
XRD patterns of the the submicrostructured (S), bimodal (B) and conventional (C) as-sprayed coatings.
The relative percentages of WC, W2C and Co6W6C phases are listed in Table 3. It is clear from the data that relative percentages of WC and W2C phases increase, with an increase of mean carbide grain size. The main reason is that higher level of decarburization of the WC occurs, with a decrease of carbide grain size. The tendence agrees reasonably well with those reported in the Usmani study15).
| As-sprayed coatings | |||
|---|---|---|---|
| S | B | C | |
| Carbide size range [μm] | 0.6–0.9 | 0.6–4.0 | 2.6–4.0 |
| Mean carbide size dWC [μm] | 0.71 ± 0.10 | 1.61 ± 0.19 | 2.93 ± 0.34 |
| Mean free path LCo [μm] | 0.38 ± 0.08 | 0.66 ± 0.11 | 1.03 ± 0.27 |
| Porosity [vol. %] | 4.91 ± 0.81 | 2.63 ± 0.72 | 2.12 ± 0.50 |
| WC [wt. %] | 64.85 | 89.38 | 92.97 |
| W2C [wt. %] | 28.62 | 8.59 | 5.46 |
| Co6W6C [wt. %] | 6.53 | 2.03 | 1.57 |
Fig. 5 shows representative SEM cross-sectional images of the three coatings. It is clear that a denser coating structure can be observed for coatings B and C compared with the coating S, in which there are no crack and pores observed along with interface of the coating and substrate. The average porosity values of the coatings, measured by image analysis, increase with a decrease in carbide size as listed in Table 3. The above tendencies are in good agreement with those reported in the Yang study22). This may account for slightly thicker cobalt layer, which melted easily and deposited between WC particles gaps. A significant amount of lamellar-cracks were observed more clearly in coating S, but less in coatings B and C as presented in Figs. 5e and 5h, respectively. It suggests that smaller relative heating area with higher particel-density leads to a worse deposition, of the powder S.
SEM images of cross-sectional view of (4a, 4b, 4c) submicrostructured S, (4d, 4e, 4f) bimodal B and (4g, 4h, 4i) conventional C coatings.
There is a dispersion of WC particles embedded in the Co-rich matrix, as shown in Figs. 5c, 5f and 5i. The range and average value of carbide size in the three coatings are also listed in Table 3. It should be noted that the carbide shapes in the coatings are more rounded compared with initial morphology in the agglomerated powders. This may account for melting/decomposition of the surface of the carbide particles during spray process, which can lead to a change in the carbide morphology. Indeed, the EDS spectrum of the bright particle, light grey particle, rod-like shaped particels and dark regions with various morphologies are believed to be the W2C, WC, γ phases and Co-rich matrix. In addition, binder mean free path increase with a increase of carbide size in the coatings (listed in Table 3), which is in agreement with the results that have been reported in the Usmani investigations of other HVOF sprayed coatings15).
3.3 Mechanical PropertiesTable 4 summarized mean microhardness and indentation fracture toughness values of the coatings. The reasults reveal the mean microhardness value increases with a decrease of carbide size, whereas the indentation fracture toughness values appears to follow the opposite trend, i.e. the fracture toughness decreases with decreasing carbide size.
| As-sprayed coatings | |||
|---|---|---|---|
| S | B | C | |
| Microhardness [HV0.3] | 1223.7 ± 42.2 | 1196.3 ± 70.6 | 1036.5 ± 112.6 |
| Fracture toughness [Mpa.m−1/2] | 3.92 ± 0.91 | 5.13 ± 0.63 | 5.66 ± 0.50 |
In general, the resulting microhardness is a balance of some contributions such as the morpholoy of matrix (crystalline/amorphous/nanocrystalline), hard phase distribution (W2C and CoxWyC content), carbide size, porosity and particels cohesion, etc23). Due to the finer WC distribution, higher W2C content and amorphous matrix instead of metllic cobalt strengthening the coating by W and C in the coating S, the microhardness of the coating S is higher than that obtained for the coating B and C. The microhardness of the coating B is higher compared with coating C, while the dispersion of value is lower. The coating B having a denser structure with the deposition of finer WC between coarse particels gaps, provides a range of values with a lower dispersion along the coating. Indeed, the indentation cracks within cross section were scarce in the X-direction (direction parallel to the substrate) presented in Fig. 6, the fracture toughness values were calculated by considering the sum of crack lengths in the Y-direction (direction perpendicular to the substrate). The porosity and splat cohesion may strongly influence the indentation fracture toughness values at higher load, a lower fracture toughness may be accounted for higher porosity and weaker splat cohesion in the coating S, which is in agreement with the results in the Usmani study15).
SEM images of the identation on the submicrostructured(6a), bimodal(6b) and conventional(6c) as-sprayed cross-sectional coatings.
The various coefficient of friction and wear loss with the load of coating S, B and C are given in Tables 5 and 6. The coefficient of friction and wear loss increase with an increase of the normal force. The coating S shows the higher coefficient of friction, while wear loss for the coating B is the lowest of all coatings. However, no clear trend is noticed in the coefficient of friction and wear loss values with reaction mean carbide size, i.e. the coating B has a lower coefficient of friction as compared to the coating S and C, whereas wear loss is higher than the coating S under a load of 50 N.
| Normal force [N] | Coefficient of friction | ||
|---|---|---|---|
| S | B | C | |
| 50 | 0.78 ± 0.09 | 0.71 ± 0.06 | 0.73 ± 0.08 |
| 150 | 0.71 ± 0.05 | 0.62 ± 0.05 | 0.69 ± 0.06 |
| Normal force [N] | Weight loss [mg] | ||
|---|---|---|---|
| S | B | C | |
| 50 | 0.34 ± 0.10 | 0.39 ± 0.12 | 0.93 ± 0.12 |
| 150 | 0.56 ± 0.19 | 0.51 ± 0.16 | 1.44 ± 0.23 |
The dependence of the coefficient of friction on the below mechanisms such as matrix material elastic deformation, adhesion, hard protrusions, asperities, ploughing proposed in the several literatures12,13,24). At the beginning stage of the test, two surfaces are brought into sliding contact, the soft ductile cobalt matrix between WC particles suffers severe deformation, then forms a thin layer at the sliding interface (called for tribofilm)16), which reduces the coefficient of friction by acting as a lubricant. There is no signs of asperity deformation in the wear track. Indeed, the coefficient of friction response is closely related to microstructure characteristics. The result of the lower coefficient of friction for the coating B and C is attributable to increased cobalt layer thickness arising from the larger carbide size. Due to presence of certain quantity of hard protrusions formed by fatigue-fracture and ploughing by wear particles wedged between these sliding surfaces (presented in Fig. 7a, grooves in the red cycle internal), the coefficient of friction for the coating S is higher than obtained for the coating C. With increasing normal force, the sliding time to the steady state shortens significantly, which leads to forming a smooth coating. Hence, the coefficient of friction for the coatings is lower under a higher load.
Sliding wear track observations of the submicrostructured (7a, 7b), bimodal (7c, 7d) and conventional (7e, 7f) as-sprayed coatings under a load of 150 N. (Figs. 7a, 7c and 7e are global view; Figs. 7b, 7d and 7f are the corresponding to enlarged view in the red box)
According to several authors15,22), it should be noted that the sliding wear resistance is related more to microhardness than to fracture toughness, i.e. wear loss increases as microhardness decreases. In contrast, the abrasion resistance of the coating appears to be related more to their fracture toughness than to their carbide size and hardness. However, other factors such as W2C content, WC grain size, dissolution, not only make the coefficient of friction change (above experiment results), but also can modify the corresponding to sliding wear resistance for a given microhardness value. Thus, the sliding wear resistance increases with increasing carbide size under a load of 50 N. The higher wear resistance of the coating S appears to be determined by higher microhardness in this case without an associated decrease in fracture toughness. However, the coating B has higher sliding wear resistance than the coating S under a larger load. The result is agreement with the conclusion obtained by Zhu25). Although the microhardness of the coating B is lower than the coating S, the lower dissolution of the finer WC particles, the higher fracture toughness and deposition of these finer carbides between the coarse carbides gaps in the coating B avoided wearing the matrix during the test.
The coatings tested under a 50 N load present very light grooves on the wear track, and then have not been further examined. Fig. 7 shows the worn surfaces of the coatings under a load of 150 N. It is clearly observed that the mechanism of surface damage of the coatings followed: propagation of cracks through the Co matrix or poorly cohesive splats, carbide fracture and pulled out. These cracks are caused by a fatigue mechanism and are initiated at the defects like porosity and detrimental phases like W2C and Co6W6C, then interlinking of subsurface cracks propagate in the binder and along the splat boundaries with a direction perpendicular to the sliding direction as shown in Figs. 7b. This may be attributed to larger shear and more brittle W2C content. Nevertheless, length and intensity of cracks within wear track for the coating C is smaller (presented in Fig. 7d), which is result of higher fracture toughness. In the coating C, the presence of the soft and ductile cobalt matrix forming a tribofilm of reattached debris, lowered the coefficient of friction, and then cobalt extrusion followed by carbide removal, or carbide fracture, but propagation of binder cracks is scarce. It is evident from wear track that there are some shallow grooves formed by ploughing in the coatings S and B. In addition, it is also interesting to mention that the wear tracks of the coatings S and B are really thin when compared with that obtained for the coating C, which can more readily give an idea of the lower damage products.
In present work, the WC-17 Co coatings were deposited by an atmospheric plasma spray process on carbon steel substrate using three agglomerated powders with different carbide size. The microstructure, phase distribution and mechanical properties of these coatings have been investigated and closely related to their sliding wear behavior.
Characterization of the coatings has revealed that porosity increases with decreasing carbide size except the coating B is lower, which can be attributed to bulk cobalt matrix and higher particle-density in the starting powders B and C. A finer carbide size distribution in the starting powder S leads to an increase in the W2C and Co6W6C phases within coating, whereas the mixing structure of finer and coarse carbide particels leads to a significantly lower degree of overall decomposition of WC.
Coating with finer carbides shows higher hardness, which it is closely related to brittle W2C content and mean carbide size (dWC). In addition, the higher indentation fracture toughness of the coating B is mainly caused by higher mean free path and larger splat cohesion.
The coefficient of friction for the coatings decrease slightly with increasing test load, and that friction coefficient of the coatings B and C is lower than the other one with an associated enlarge in mean binder free path. The coating B and S tested at a higher load have higher sliding wear resistance, of which predominant material removal mechanisms summarizes below: propagation of cracks through the Co matrix or poorly cohesive splats, carbide fracture and pulled out. However, Co-matrix extrusion followed by carbide removal, or carbide fracture is the predominant material removal mechanisms of the coating C.
This study was financially supported by Special Fund of BGRIMM (No. YJZ201404).
