- Email: [email protected]
Effect of W dissolution in NiCrBSi–WC and NiBSi–WC arc sprayed coatings on wear behaviors
Wear 317 (2014) 194–200
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Effect of W dissolution in NiCrBSi–WC and NiBSi–WC arc sprayed coatings on wear behaviors Panadda Sheppard n, Hathaipat Koiprasert National Metals and Materials Technology Center, Pathum Thani 12120, Thailand
art ic l e i nf o
a b s t r a c t
Article history: Received 23 January 2014 Received in revised form 2 June 2014 Accepted 6 June 2014 Available online 16 June 2014
This work concerns the arc-sprayed NiCrBSi–WC and NiBSi–WC coatings produced from cored-wires. Porosity and WC/W2C contents were determined from the as-sprayed coatings. It was suggested that the dissolution of W into the Ni-rich matrix, which was more pronounced in NiBSi–WC coating and resulting in the thermal expansion coefficient of the matrix to reduce, caused this coating to have lower thermal stress from fabrication. This resulted in a lower amount of WC/W2C detachment in NiBSi–WC coating. The dissolution of W however has an adverse effect of reducing the WC/W2C contents in the coating, which reduces its hardness. Wear testing revealed that, even though the NiCrBSi–WC coating contained higher WC/W2C content than the NiBSi–WC coating due to lower W dissolution, its performance was inferior to the latter. In the dry sliding wear test, the problem of WC/ W2C detachment on the contacting surface became exacerbated in the NiCrBSi–WC coating, leaving craters on the wear surface. In the three-body abrasive wear test, there was much less WC/W2C detachment. However the NiBSi–WC coating continued to out-perform the NiCrBSi–WC coating, suggesting that the higher W dissolution into the Ni-rich matrix has a major role in increasing the abrasive resistance of the coating. & 2014 Elsevier B.V. All rights reserved.
Keywords: Sliding wear Three-body abrasion Cermets Thermal spray coatings Tungsten dissolution
1. Introduction Carbide-reinforced NiCrBSi and NiBSi as metal matrix composites (MMCs) have been used in many applications such as oil drilling equipments, agricultural machineries, piston rings, rollers in steel making and wire drawing rolls, in order to combat mechanical degradations such as sliding wear, abrasive wear and rolling contact fatigue [1–5]. WC is one of the most commonly used hard reinforced particles due to its high hardness and toughness and its ability to be wetted easily by many molten metals [6–8], hence producing a well bonded MMC. Electric arc spraying (EAS), also known as metallization, is a thermal spraying technique used in depositing a thick metallic coating from wire form. It is a simple and cost-effective process and is popularly employed in surface refurbishments and maintenance sectors. The principle of EAS involves two electrically conductive wires fed into a common arc point, creating a temperature of up to the melting temperature of the wire material at which melting occurs. The molten material is continuously atomized and the droplets are accelerated toward the substrate by a
n
Corresponding author. Tel.: þ 66 25646500; fax: þ 66 25646401. E-mail address: [email protected] (P. Sheppard).
http://dx.doi.org/10.1016/j.wear.2014.06.008 0043-1648/& 2014 Elsevier B.V. All rights reserved.
compressed air jet [9]. As well as a solid metallic wire, cored wires where non-conductive powders are contained inside a metallic tube can also be used. Material properties that govern its wear performance, apart from its intrinsic properties such as friction coefficient, hardness and fracture toughness, also include external factors such as lubrication, temperature, environment and the material compatibility with the counter surfaces. In the case of carbide-reinforced NiCrBSi and NiBSi coatings, factors such as coating adhesion strength to the substrate, type of carbides, and carbide content [10,11] also play a major role in enhancing the wear resistance of a component. These mentioned properties have a strong relationship to the microstructure of the material. Previous work has found that W can dissolve into the NiCrBSi– WC and NiBSi–WC EAS coatings during spraying, forming solidsolutions of NiCrW and NiW at room temperature. As a consequence, the carbide contents were lowered in both coatings [12]. These occurrences undoubtedly affect the physical and mechanical properties of the coatings, and in particular, the tribological properties where the coating's main applications lie. The objective of this work is to study the effect of the W dissolution in NiCrBSi–WC and NiBSi–WC coatings produced via the EAS process on the physical and mechanical properties of the coatings.
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
2. Experimental procedure
195
Table 1 Chemical composition of sprayed wires.
2.1. Materials
Cored wire Metallic elements (totaled to 50 wt%) 3
Stainless steel 304 coupons of dimension 5 25.4 60 mm and 10 10 5 mm3 were used as substrates for sliding wear test specimens and abrasive wear test specimens, respectively. All substrate specimens were grit blasted using 740 μm Al2O3 in order to achieve the surface roughness of 8–10 μm Ra. The high surface roughness is essential to ensure adequate bonding of the coatings. The degree of surface roughness required depends on the coating material, the process and the service condition of the coating. The minimum roughness of 8 μm Ra was chosen according to an industrial practice of producing an arc sprayed wear-resistant coating of less than 300 μm thickness. Higher degree of the substrate roughness will result in higher bonding strength of the coating but will also increase the roughness of the finished coating. The specimens were then ultrasonically cleaned and dried, and were ready for the EAS coating process. Two groups of coating samples were produced using EAS from 2 types of cored wire. The wire compositions are shown in Table 1. 2.2. Electric arc spraying (EAS) The specimens were spray-coated using Sulzer Metco SmartArc to achieve a coating thickness of 250–300 mm. For wear applications, thicker coating is generally more desirable to allow for a longer lifetime. However the maximum coating thickness is partly controlled by the loss of the outer surface roughness as the coating builds up and also the reduction in its bond strength as the coating becomes thicker. Hence the 250–300 mm thickness was chosen as a moderate thickness that still allow for a usable wear-resistant coating. The optimized spraying parameters and the sprayed materials for two groups of sample, a and b, are shown in Table 2. The optimization was done by varying the spray distance and the arc voltage, which subsequently controls the arc current and the wire feed rate. The chosen condition produced a coating with the lowest percentages of porosity and oxide content. 2.3. Physical properties The samples were cross-sectioned and polished to reveal the coating microstructures. JEOL JSM5410 scanning electron microscope (SEM) and energy dispersive X-ray spectroscopy (EDS) were used to characterize the structures and phase compositions of the coatings. The percentage of cracks and porosities of the coatings were determined using image analysis [13] (Image Pro Plus version 5.1) on optical images of the coating cross sections. Hardness testing in kg/mm2 using Vickers microhardness at 0.4 kg loading and Brinell hardness at 187.5 kg loading were also utilized. An averaged Vickers hardness number was obtained from 30 indentations across the polished cross-section of the coating in order to represent the heterogeneous nature of the coating. An averaged Brinell hardness number was obtained from 10 indentations on the polished surface of the coating to represent the effective hardness of the coated specimen. Due to the gradual variation of chemical composition in coating splats as a result of W dissolution [12], a clear distinction does not occur between WC and the Ni-rich phases. In this work, etching was therefore employed in order to identify the carbide phases. The chemical mixture was 1:1 of 70% conc. HNO3:H2O by volume. The polished planar and cross-section samples were exposed to the mixture for up to 5 min at 75 1C. Splats exhibiting no sign of corrosion attack were determined as carbides. Image analysis was then used to measure their area percentages.
a b
Ni
Cr
B
Si
C
Bal. Bal.
10–12 –
1.7–2.0 1.7–2.0
4–5 4–5
0.4 0.4
Fused WC/W2C 50 wt%
þ100% þ100%
Fused tungsten carbide consists of 78–80% W2C and 20–22%WC by weight. WC/W2C particle size is 1007 40 mm.
Table 2 Optimized spraying parameters. Sample
a
b
Cored wire Air pressure (kN/m2) Arc voltage (V) Arc current (A) Feed rate (kg/s) Spray distance (mm)
a 345 32–35 150 0.0028 127
b
2.4. Tribological testing 2.4.1. Dry sliding wear The as-sprayed samples were polished in the planar direction to achieve the surface roughness of 0.7 mm Ra. Linear reciprocating ball-on-flat dry sliding wear tests were performed on Micro Tribometer UMT 2 test machine (Bruker Instrument, USA). The test setup and parameters adhered to ASTM G 133-95 procedure A using 25 N normal force, 10 mm stroke length, 5 Hz oscillating frequency and 1000 s test duration in an unlubricated condition at room temperature [14]. The pin tip radius of 4.76 mm however was not adopted. Instead a through-hardened high carbon steel ball (AISI 52100) of 6.3 mm diameter and hardness of 60–64 Rockwell C was used as the contacting spherical surface. Three tests were conducted for each group of specimen. The wear damages were reported as wear depths of the specimens. In order to obtain the wear depth value, the tested specimen was crosssectioned and examined through an optical microscope. The lowest point on the wear track was taken as the wear depth. 2.4.2. Abrasive wear The as-sprayed samples were polished to 0.7 mm Ra in the planar direction. A wheel grinder was employed for the abrasive test, using a wheel speed of 30 rpm. The specimens were under a 20 N load and moved in the opposite direction to the wheel at 60 rpm. The grinding wheel was padded with a polishing cloth. A continuous feed of 5 μm alumina suspension (Buehler Micropolish II Deagglomerated) was set at 2 ml/min. Three tests were conducted for each group of specimen. Weight changes of the specimens were used to evaluate the abrasive wear performance.
3. Results and discussion 3.1. Coating microstructures Cross sections of the coatings were revealed using the backscattered SEM, see Fig. 1. Both coatings display good integrity. No large cracks were observed. There are however some small globular pores and intergranular pores observed in coatings a and b. In both samples, WC/W2C (light colored phase) scatters throughout the Ni-rich matrix (dark colored phase). The Ni-rich phase varies significantly in its chemical composition, and hence
196
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
WC/W2C enveloped in Ni-rich phase NiCr-rich phase
Ni-rich phase
Ni,Cr-oxide
Fig. 1. Micrographs of (a) coating a, and (b) coating b.
Fig. 2. Schematic diagram showing the physical actions of sprayed particles during the EAS process.
the shading in Fig. 1. This was previously shown to be as a result of W dissolution into the Ni-rich phase [12]. Oxides of Ni,Cr and Ni are found in coating a and b, respectively. These oxides formed when the sprayed particles were in flight. Where WC/W2C bonded with the Ni-rich phases, the interfaces are intimate with no porosity or oxides present. This is because WC and W2C are readily wetted by molten Ni and NiCr. During EAS spraying, the wetting occurred as soon as the Ni and NiCr sheaths began to melt at the wire tips, protecting the Ni–WC/W2C interface from further oxidation. But at the same time, W dissolution can occur easier at this interface. The exposed Ni-rich surfaces however were in contact with oxygen at high temperature, resulting in surface oxide layers around the particles, which became incorporated into the coatings, see Figs. 1 and 2. The amounts of oxide in both coatings show large variations from one location to another with most regions indicating the value of less than 0.1% making quantitative measurement not reliable. Small cracks occurred throughout both coatings. These cracks can be categorized as intergranular and lateral cracks, with the intergranular crack being much more frequent, see Fig. 3. The short lateral cracks can occasionally be observed. The cracks are caused by thermal stress arising from rapid solidification and thermal expansion mismatch between WC/W2C and Ni-rich phases. This is followed by stress relaxation by cracking during cooling from the spraying temperature [15]. Previous works have shown that the presence of Cr in Ni and its alloys has an effect on
Intergranular crack
Lateral crack
Fig. 3. Cracks in WC/W2C in sample b.
their thermal expansion coefficients. Its effect however is small when compared to that of the W [16,17]. W dissolves into Ni-rich phases during the EAS coating process aided by the readily wetting of Ni. The amount of dissolved W depends on the existing amount of Cr in the Ni phase. When there is no Cr, more W can diffuse into Ni. But as Cr increases, a lower amount of W can dissolve into solid solution with Ni–Cr. This is in agreement with the Cr–Ni–W equilibrium phase diagram. At 1200 1C, the solubility limit of W in Ni is as much as 40 wt%.
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
197
Table 3 Thermal expansion coefficients of Ni alloys with various chemical compositions, calculated from a model proposed by Dosovitskiy et al. (2009), and of WC [17,19,20]. Chemical composition (wt%)
Thermal expansion coefficient ( 10 6 1C 1) at 20–800 1C
Ni
Cr
W
WC
Bal. Bal. Bal. Bal. Bal. – –
– 20 – – 20.0 – –
– – 5.4 20.0 5.0 100 –
– – – – – – 100
Intergranular crack Detaching carbide
15.9 15.4 14.6 12.6 14.7 4.5–5.0 4.8
When 20 wt% of Cr is in solution with Ni, however, the W solubility limit in NiCr reduces to lower than 18 wt% [18]. Therefore, there was more W diffused into coating b matrix, making the major phase Ni–W, than there was W diffused into coating a matrix. As much as 20 wt% of W can often be seen in the Ni-rich phase in coating b, occasionally reaching as high as 50%. While only 5 wt% of W is commonly observed in NiCr-rich phase in coating a. W is a refractory metal with the lowest thermal expansion coefficient (TEC) amongst pure metal of about 4.5 10 6 1C 1 at room temperature. When W is added to other metal alloys, it can alter their TECs. The thermal expansion of NiW, NiCr, and NiCrW alloys has been measured by quartz dilatometry for the 20–800 1C temperature range by Dosovitskiy et al. [17]. It was found that substitution of Ni by W leads to a considerable decrease in TEC, while Cr has only a small influence on the TEC of the alloy. The prediction model from this work was utilized in creating Table 3 for the purpose of comparing the TECs of the Ni-rich splats to that of the WC particles [19,20]. In sample b, where it is common to find as much as 20 wt% or more of W in the Ni splats, which corresponds to a drop of TEC to below 12.6 10 6 1C 1. By contrast, it is infrequent that a NiCr splat containing more than 5 wt% of W is detected in sample a. A Ni-rich splat containing 20 wt% Cr and 5 wt% W is corresponding to the TEC of approximately 14.7 10 6 1C 1. Hence, upon rapid cooling from spraying temperature, coating a (NiCrW) suffered higher thermal stress than coating b (NiW) as a result of the higher contraction as well as the thermal expansion mismatch with the WC/W2C particles. The higher thermal stress leads to the weakening of the inter-splat bonds, thus allowing easier propagation of intergranular cracks. The image analysis technique used in measuring the porosity level of the coatings detects both pores and cracks as porosity. The lateral cracks are often very small and so are not picked up as porosity. The intergranular cracks due to the thermal contraction of the coatings, on the other hand, are large enough to be included in the total values of the porosity. Apart from the cracks, there were also small globular pores as a result of incomplete contact between successive splats [13] and voids due to the particle detachment, see Fig. 4. Table 4 shows the levels of porosity in the coatings. The value distributions are large due to the heterogeneous nature of the coatings. The higher percentage of porosity in coating a is a result of a larger amount of the intergranular crack. Furthermore, the stress from cutting and polishing of the sample cross sections, together with the weaker inter-splat bonding due to large thermal stress obtained during coating fabrication in coating a, inevitably dislodged some particles, leaving behind characteristic rounded craters, which were detected as pores, see Fig. 5. In coating b, the thermal stress from coating fabrication also caused the intergranular cracks. The lower thermal expansion mismatch between
Intergranular crack
Carbide
Fig. 4. Intergranular cracks in (a) sample a and (b) sample b.
Table 4 Percentage porosities and hardness values of coatings.
Sample a Sample b
% Porosity
WC/W2C content (vol%)
Microhardness (Hv0.4)
Brinell hardness (Hb)
2.02 7 0.70 0.90 7 0.50
24.37 2.8 17.3 71.9
5727 162 381 7 167
1657 6.4 1517 2.7
Crater left by carbide Fig. 5. Detachment of carbide particle in sample a.
splats can result in lower stress at the intergranular interface and lower carbide detachment, see Fig. 6. Hence, the total porosity is lower in sample b than in sample a.
198
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
Intergranular crack
Crater from carbide detachment
Deformed carbide
Fig. 6. Debonding of carbide particle in sample b. Fig. 8. Sliding wear track of sample a. Table 5 Coating compositions. (2) Calculated %WC (vol%), (3) Measured %WC (1) Coating (vol%), using Image if all W in composition analysis technique (converted to vol% (1) is in WC form from wt%) measured using EDS Ni Sample a 58 Sample b 78.7
Cr
W
19.3 –
22.7 21.3
27.4 26.1
24.3 17.2
Results calculated using density of Ni¼ 8.9 g/cm3, density of Cr¼ 7.1 g/cm3, density of W ¼ 19.25 g/cm3 and density of WC ¼ 15.8 g/cm3.
Crater from carbide detachment
Fig. 9. Sliding wear track of sample b.
with the starting cored wire materials, it is clear that a large percentage of the carbide was loss during spraying. It is not unusual in the thermal spraying process to lose certain phases more than others, particularly if the variation in size and density are large as in this case. In column (3), however, when an image analysis technique was used to determine the volume percent of the carbide phase, it was found that a large decline occurs in sample b. This finding supports earlier results which proposed that the carbide phase decomposed and W dissolved into the Ni-rich phase and that a much larger extent of dissolution occurred in sample b than in sample a. Fig. 7. Wear depths and coefficient of frictions determined by dry sliding wear test.
3.3. Coating hardness 3.2. WC contents WC/W2C content varies from one area of the coating to another due to the heterogeneous structure of the EAS coating. Table 4 shows the WC/W2C contents of the coatings. The average values indicate that, by percentage volume, there is a larger amount of carbides in sample a than there is in sample b. Given that the starting cored wire contains 50% by weight or about 35–36% by volume of WC/W2C, the experimental data of 24.3 and 17.3 vol% for sample a and b, respectively, may appear on a low side. Quantitative EDS was employed to measure the total coating compositions at 600 magnification with the light elements (in this case B, C and O) omitted. The results are presented in Table 5 column (1). In column (2), it was then assumed that all W from column (1) is in the form of WC. In which case, C was manually added to the calculation, resulting in the WC volume percentage of 27.4 and 26.1 for coating a and b, respectively. When compared
Table 4 reports the average hardness values of the coatings. The dissolution of W results in large variation of the hardness throughout the coating. This can be observed when Vickers microhardness is performed. The results show a very large distribution. Indentation size effect (ISE) can be an issue in the utilization of the microhardness test. In this test however the indent size of greater than 10 mm width is comparable to the thickness of the splat rendering ISE less dominant [21]. A better representative of the hardness can be obtained using the macrohardness Brinell method on the surface of the coating. However the indent size in this case is large in comparison to the coating thickness. Hence the hardness of the underlying substrate affects the measurement. Nevertheless Brinell testing takes into account the defects in the coating and can give a better representative of the effective hardness than the Vickers test. The lower Brinell hardness in coating b is due to a larger amount of W being dissolved in the Ni matrix in coating b than there is in the matrix
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
Weight Change (g)
Time (min.) 20
0 -0.002
0
40
60
80
10
20
-0.004 -0.006
sample a
-0.008
sample b
-0.01
Fig. 10. Result from abrasive wear test.
Crater
199
0.15 70.01 mm and 0.12 70.01 mm respectively. The low distributions in the test results indicate the good repeatability of the test. The reason for the difference in the wear depth is related to the intergranular cracks and the detachment of WC/W2C particles. In sample a, where carbide detachment can occur more easily, even though the coating started off with higher amount of carbide particles the stress induced by the sliding action caused the intergranular cracks to propagate and WC/W2C to be pulled-out, allowing the softer matrix to wear more rapidly. Figs. 8 and 9 reveal the wear tracks on the two coatings. The sliding direction is from left to right in both pictures. Both samples show little deformation on the wear tracks. There are, however, a large number of craters on the surface of sample a. These craters are the result of WC/W2C detachments during the sliding wear testing. There are fewer surface craters present on the surface of sample b due to better bonding between splats as a result of lower thermal stress during fabrication. 3.5. Abrasive wear
Carbide
Fig. 11. Abrasive wear surface of sample a.
Ni-W solid solution
Carbide Crater
Fig. 12. Abrasive wear surface of sample b.
Table 6 Average measurement of WC/W2C pull-out after the abrasive wear test. WC/W2C pull-out (vol%) Sample a Sample b
1.72 71.04 2.66 71.03
The results from the three-body abrasive wear testing in Fig. 10 show that sample b again performed better than sample a. The wear surfaces in Figs. 11 and 12 however demonstrate that the reason for this outcome is markedly different from the sliding wear test. In the abrasive case, there are some WC/W2C detachments in both coatings, see Table 6. The extent of the carbide detachment is significantly lower in sample a, when comparing the surfaces after the abrasive test to those after the sliding test. The difference in the carbide pull-out is not as extreme in sample b. This is because the cutting action of the loose alumina abrasives put less strain on the coatings, rendering the weak carbide-matrix bonding not as important as in the sliding case, where deformation and seizure between the ball and the coatings promoted the intergranular crack propagation. If the WC/W2C detachment is minimal then sample a, which contains a larger proportion of carbide and possesses a higher hardness, should outperform sample b in the abrasive wear test. With this contradiction, the focus was turned to the Ni-rich matrix. Cr was originally added to Ni in the cored wire used to produce coating a with an intention that the presence of Cr would promote the formation of hard phases such as Cr3C2 and Cr7C3 which can improve the wear resistance of the coating. However, the short spraying process does not allow for the Cr carbide formation and there was no significant amount of Cr carbide detected in this study. Thus Cr does not significantly contribute to the coating hardness via the formation of Cr carbides. Tungsten as an alloying element in Ni on the other hand has been shown in previous works to have a positive influence on the hardness and fracture toughness, thus improving the abrasive resistance of the material [22,23]. The wear surface of sample b in Fig. 12 shows that the Ni–W phase possesses superior abrasion resistance to the Ni phase, presumably as a result of solid solution strengthening by W, causing the wear surface to be uneven. This Ni–W is thus believed to play a major role in providing the abrasive wear resistance for the coating.
4. Conclusions in coating a. This inevitably results in a greater loss of WC/W2C phase in coating b, thus resulting in the lower hardness. 3.4. Dry sliding wear
For NiCrBSi–WC and NiBSi–WC coatings produced from cored wires using the EAS technique, the results can be summarized as follows:
Sliding wear tests were performed using steel balls on the polished surfaces of the coatings. The results in Fig. 7 show that sample a exhibits higher wear depth than sample b, at
1. W from the WC/W2C dissolved into the Ni-rich matrix during spraying. There was more W content in NiBSi (sample b matrix) than there was in NiCrBSi (sample a matrix), resulting in a
200
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
lower thermal expansion mismatch in NiBSi, hence lower thermal stress from the coating fabrication. The thermal stress in the coatings may propagate intergranular cracks, resulting in more WC/W2C being debonded in the NiCrBSi–WC coating (sample a) where less W was detected in the Ni-rich matrix. 2. The NiBSi–WC coating contained a lower amount of WC/W2C phase than the NiCrBSi–WC coating due to extensive dissolution of W into the Ni-rich matrix. 3. Sliding action during the dry sliding wear test caused more WC/W2C to be detached from the NiCrBSi–WC coating, resulting in higher wear depth than NiBSi–WC coating. 4. Abrasive wear testing showed that the NiBSi–WC coating had slightly higher abrasive resistance than the NiCrBSi–WC coating, even though it contained a lower percentage of carbide than the latter. This is due to the presence of the extensive Ni–W phase.
References [1] S.P. Lu, O.Y. Kwon, Y. Guo, Wear behavior of brazed WC/NiCrBSi(Co) composite coatings, Wear 254 (5–6) (2003) 421–428. [2] S. Stewart, R. Ahmed, T. Itsukaichi, Contact fatigue failure evaluation of post-treated WC–NiCrBSi functionally graded thermal spray coatings, Wear 257 (9–10) (2004) 962–983. [3] A. Martín, J. Rodríguez, J.E. Fernández, R. Vijande, Sliding wear behaviour of plasma sprayed WC–NiCrBSi coatings at different temperatures, Wear 251 (1–2) (2001) 1017–1022. [4] N.Y. Sari, M. Yilmaz, Improvement of wear resistance of wire drawing rolls with Cr–Ni–B–Siþ WC thermal spraying powders, Surf. Coat. Technol. 202 (13) (2008) 3136–3141. [5] L.C. Betancourt-Dougherty, R.W. Smith, Effects of load and sliding speed on the wear behaviour of plasma sprayed TiC–NiCrBSi coatings, Wear 217 (1) (1998) 147–154. [6] J. Przybylowicz, J. Kusinski, Structure of laser cladded tungsten carbide composite coatings, J. Mater. Process Technol. 109 (1–2) (2001) 154–160.
[7] J. He, J.M. Schoenung, A review on nanostructured WC–Co coatings, Surf. Coat. Technol. 157 (1) (2002) 72–79. [8] T.Y. Cho, J.H. Yoon, K.S. Kim, K.O. Song, Y.K. Joo, W. Fang, S.H. Zhang, S.J. Youn, H.G. Chun, S.Y. Hwang, A study on HVOF coatings of micron and nano WC–Co powders, Surf. Coat. Technol. 202 (22–23) (2008) 5556–5559. [9] S.D. Cramer, B.S. Covino Jr. (Eds.), ASM Handbook Vol. 13A, Corrosion: Fundamentals, Testing, and Protection, ASM International, Ohio, 804. [10] P. Wu, C.Z. Zhou, X.N. Tang, Microstructural characterization and wear behavior of laser cladded nickel-based and tungsten carbide composite coatings, Surf. Coat. Technol. 166 (2003) 84–88. [11] K.V. Acker, D. Vanhoyweghen, R. Persoons, J. Vangrunderbeek, Influence of tungsten carbide particle size and distribution on the wear resistance of laser clad WC/Ni coatings, Wear 258 (2005) 194–202. [12] P. Niranatlumpong, H. Koiprasert, Phase transformation of NiCrBSi–WC and NiBSi–WC arc sprayed coatings, Surf. Coat. Technol. 206 (2011) 440–445. [13] S. Deshpande, A. Kulkarni, S. Sampath, H. Herman, Application of image analysis for characterization of porosity in thermal spray coatings and correlation with small angle neutron scattering, Surf. Coat. Technol. 187 (1) (2004) 6–16. [14] Standard test method for linearly reciprocating ball-on-flat sliding wear, G 133-95, Annual Book of ASTM Standards, 03.02, ASTM, 1995, pp. 529–536. [15] H. Herman, Plasma sprayed coatings, Sci. Am. 259 (3) (1988) 112–117. [16] F. Xiao, R. Yang, L. Fang, C. Zhang, Solidification shrinkage of Ni–Cr alloys, Mater. Sci. Eng. B 132 (2006) 193–196. [17] G.A. Dosovitskiy, S.V. Samoilenkov, A.R. Kaul, D.P. Rodionov, Thermal expansion of Ni–W, Ni–Cr, and Ni–Cr–W alloys between room temperature and 800 1C, Int. J. Thermophys. 30 (2009) 1931–1937. [18] K.P. Gupta, The Cr–Ni–W system, Phase Diagrams of Ternary Nickel Alloys, Monograph Series on Alloy Phase Diagrams, The Indian Institute of Metals, Calcutta (1990) 111–133. [19] H. Doi, Y. Fujiwara, K. Miyake, Y. Oosawa, A systematic investigation of elastic moduli of WC–Co alloys, Metall. Trans. 1 (5) (1970) 1417–1425. [20] E. Lassner, W.D. Schubert, Tungsten: Properties, Chemistry, Technology of the Element, Alloys and Chemical Compounds, Kluwer Academic, Plenum Publishers, New York (1999) 31. [21] I. Manika, J. Maniks, Size effects in micro- and nanoscale indentation, Acta Mater. 54 (2006) 2049–2056. [22] C.A. Schuh, T.G. Nieh, H. Iwasaki, The effect of solid solution W additions on the mechanical properties of nanocrystalline Ni, Acta Mater. 51 (2) (2003) 431–443. [23] K.T. Kim, J.H. Lim, J.H. Kim, J. Joo, W. Nah, B.K. Ji, B.H. Jun, C.J. Kim, G.W. Hong, Development of cube-textured Ni–W alloy substrates for YBCO-coated conductor, Phys. C: Supercond. 412–414 (2) (2004) 859–863.
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Effect of W dissolution in NiCrBSi–WC and NiBSi–WC arc sprayed coatings on wear behaviors Panadda Sheppard n, Hathaipat Koiprasert National Metals and Materials Technology Center, Pathum Thani 12120, Thailand
art ic l e i nf o
a b s t r a c t
Article history: Received 23 January 2014 Received in revised form 2 June 2014 Accepted 6 June 2014 Available online 16 June 2014
This work concerns the arc-sprayed NiCrBSi–WC and NiBSi–WC coatings produced from cored-wires. Porosity and WC/W2C contents were determined from the as-sprayed coatings. It was suggested that the dissolution of W into the Ni-rich matrix, which was more pronounced in NiBSi–WC coating and resulting in the thermal expansion coefficient of the matrix to reduce, caused this coating to have lower thermal stress from fabrication. This resulted in a lower amount of WC/W2C detachment in NiBSi–WC coating. The dissolution of W however has an adverse effect of reducing the WC/W2C contents in the coating, which reduces its hardness. Wear testing revealed that, even though the NiCrBSi–WC coating contained higher WC/W2C content than the NiBSi–WC coating due to lower W dissolution, its performance was inferior to the latter. In the dry sliding wear test, the problem of WC/ W2C detachment on the contacting surface became exacerbated in the NiCrBSi–WC coating, leaving craters on the wear surface. In the three-body abrasive wear test, there was much less WC/W2C detachment. However the NiBSi–WC coating continued to out-perform the NiCrBSi–WC coating, suggesting that the higher W dissolution into the Ni-rich matrix has a major role in increasing the abrasive resistance of the coating. & 2014 Elsevier B.V. All rights reserved.
Keywords: Sliding wear Three-body abrasion Cermets Thermal spray coatings Tungsten dissolution
1. Introduction Carbide-reinforced NiCrBSi and NiBSi as metal matrix composites (MMCs) have been used in many applications such as oil drilling equipments, agricultural machineries, piston rings, rollers in steel making and wire drawing rolls, in order to combat mechanical degradations such as sliding wear, abrasive wear and rolling contact fatigue [1–5]. WC is one of the most commonly used hard reinforced particles due to its high hardness and toughness and its ability to be wetted easily by many molten metals [6–8], hence producing a well bonded MMC. Electric arc spraying (EAS), also known as metallization, is a thermal spraying technique used in depositing a thick metallic coating from wire form. It is a simple and cost-effective process and is popularly employed in surface refurbishments and maintenance sectors. The principle of EAS involves two electrically conductive wires fed into a common arc point, creating a temperature of up to the melting temperature of the wire material at which melting occurs. The molten material is continuously atomized and the droplets are accelerated toward the substrate by a
n
Corresponding author. Tel.: þ 66 25646500; fax: þ 66 25646401. E-mail address: [email protected] (P. Sheppard).
http://dx.doi.org/10.1016/j.wear.2014.06.008 0043-1648/& 2014 Elsevier B.V. All rights reserved.
compressed air jet [9]. As well as a solid metallic wire, cored wires where non-conductive powders are contained inside a metallic tube can also be used. Material properties that govern its wear performance, apart from its intrinsic properties such as friction coefficient, hardness and fracture toughness, also include external factors such as lubrication, temperature, environment and the material compatibility with the counter surfaces. In the case of carbide-reinforced NiCrBSi and NiBSi coatings, factors such as coating adhesion strength to the substrate, type of carbides, and carbide content [10,11] also play a major role in enhancing the wear resistance of a component. These mentioned properties have a strong relationship to the microstructure of the material. Previous work has found that W can dissolve into the NiCrBSi– WC and NiBSi–WC EAS coatings during spraying, forming solidsolutions of NiCrW and NiW at room temperature. As a consequence, the carbide contents were lowered in both coatings [12]. These occurrences undoubtedly affect the physical and mechanical properties of the coatings, and in particular, the tribological properties where the coating's main applications lie. The objective of this work is to study the effect of the W dissolution in NiCrBSi–WC and NiBSi–WC coatings produced via the EAS process on the physical and mechanical properties of the coatings.
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
2. Experimental procedure
195
Table 1 Chemical composition of sprayed wires.
2.1. Materials
Cored wire Metallic elements (totaled to 50 wt%) 3
Stainless steel 304 coupons of dimension 5 25.4 60 mm and 10 10 5 mm3 were used as substrates for sliding wear test specimens and abrasive wear test specimens, respectively. All substrate specimens were grit blasted using 740 μm Al2O3 in order to achieve the surface roughness of 8–10 μm Ra. The high surface roughness is essential to ensure adequate bonding of the coatings. The degree of surface roughness required depends on the coating material, the process and the service condition of the coating. The minimum roughness of 8 μm Ra was chosen according to an industrial practice of producing an arc sprayed wear-resistant coating of less than 300 μm thickness. Higher degree of the substrate roughness will result in higher bonding strength of the coating but will also increase the roughness of the finished coating. The specimens were then ultrasonically cleaned and dried, and were ready for the EAS coating process. Two groups of coating samples were produced using EAS from 2 types of cored wire. The wire compositions are shown in Table 1. 2.2. Electric arc spraying (EAS) The specimens were spray-coated using Sulzer Metco SmartArc to achieve a coating thickness of 250–300 mm. For wear applications, thicker coating is generally more desirable to allow for a longer lifetime. However the maximum coating thickness is partly controlled by the loss of the outer surface roughness as the coating builds up and also the reduction in its bond strength as the coating becomes thicker. Hence the 250–300 mm thickness was chosen as a moderate thickness that still allow for a usable wear-resistant coating. The optimized spraying parameters and the sprayed materials for two groups of sample, a and b, are shown in Table 2. The optimization was done by varying the spray distance and the arc voltage, which subsequently controls the arc current and the wire feed rate. The chosen condition produced a coating with the lowest percentages of porosity and oxide content. 2.3. Physical properties The samples were cross-sectioned and polished to reveal the coating microstructures. JEOL JSM5410 scanning electron microscope (SEM) and energy dispersive X-ray spectroscopy (EDS) were used to characterize the structures and phase compositions of the coatings. The percentage of cracks and porosities of the coatings were determined using image analysis [13] (Image Pro Plus version 5.1) on optical images of the coating cross sections. Hardness testing in kg/mm2 using Vickers microhardness at 0.4 kg loading and Brinell hardness at 187.5 kg loading were also utilized. An averaged Vickers hardness number was obtained from 30 indentations across the polished cross-section of the coating in order to represent the heterogeneous nature of the coating. An averaged Brinell hardness number was obtained from 10 indentations on the polished surface of the coating to represent the effective hardness of the coated specimen. Due to the gradual variation of chemical composition in coating splats as a result of W dissolution [12], a clear distinction does not occur between WC and the Ni-rich phases. In this work, etching was therefore employed in order to identify the carbide phases. The chemical mixture was 1:1 of 70% conc. HNO3:H2O by volume. The polished planar and cross-section samples were exposed to the mixture for up to 5 min at 75 1C. Splats exhibiting no sign of corrosion attack were determined as carbides. Image analysis was then used to measure their area percentages.
a b
Ni
Cr
B
Si
C
Bal. Bal.
10–12 –
1.7–2.0 1.7–2.0
4–5 4–5
0.4 0.4
Fused WC/W2C 50 wt%
þ100% þ100%
Fused tungsten carbide consists of 78–80% W2C and 20–22%WC by weight. WC/W2C particle size is 1007 40 mm.
Table 2 Optimized spraying parameters. Sample
a
b
Cored wire Air pressure (kN/m2) Arc voltage (V) Arc current (A) Feed rate (kg/s) Spray distance (mm)
a 345 32–35 150 0.0028 127
b
2.4. Tribological testing 2.4.1. Dry sliding wear The as-sprayed samples were polished in the planar direction to achieve the surface roughness of 0.7 mm Ra. Linear reciprocating ball-on-flat dry sliding wear tests were performed on Micro Tribometer UMT 2 test machine (Bruker Instrument, USA). The test setup and parameters adhered to ASTM G 133-95 procedure A using 25 N normal force, 10 mm stroke length, 5 Hz oscillating frequency and 1000 s test duration in an unlubricated condition at room temperature [14]. The pin tip radius of 4.76 mm however was not adopted. Instead a through-hardened high carbon steel ball (AISI 52100) of 6.3 mm diameter and hardness of 60–64 Rockwell C was used as the contacting spherical surface. Three tests were conducted for each group of specimen. The wear damages were reported as wear depths of the specimens. In order to obtain the wear depth value, the tested specimen was crosssectioned and examined through an optical microscope. The lowest point on the wear track was taken as the wear depth. 2.4.2. Abrasive wear The as-sprayed samples were polished to 0.7 mm Ra in the planar direction. A wheel grinder was employed for the abrasive test, using a wheel speed of 30 rpm. The specimens were under a 20 N load and moved in the opposite direction to the wheel at 60 rpm. The grinding wheel was padded with a polishing cloth. A continuous feed of 5 μm alumina suspension (Buehler Micropolish II Deagglomerated) was set at 2 ml/min. Three tests were conducted for each group of specimen. Weight changes of the specimens were used to evaluate the abrasive wear performance.
3. Results and discussion 3.1. Coating microstructures Cross sections of the coatings were revealed using the backscattered SEM, see Fig. 1. Both coatings display good integrity. No large cracks were observed. There are however some small globular pores and intergranular pores observed in coatings a and b. In both samples, WC/W2C (light colored phase) scatters throughout the Ni-rich matrix (dark colored phase). The Ni-rich phase varies significantly in its chemical composition, and hence
196
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
WC/W2C enveloped in Ni-rich phase NiCr-rich phase
Ni-rich phase
Ni,Cr-oxide
Fig. 1. Micrographs of (a) coating a, and (b) coating b.
Fig. 2. Schematic diagram showing the physical actions of sprayed particles during the EAS process.
the shading in Fig. 1. This was previously shown to be as a result of W dissolution into the Ni-rich phase [12]. Oxides of Ni,Cr and Ni are found in coating a and b, respectively. These oxides formed when the sprayed particles were in flight. Where WC/W2C bonded with the Ni-rich phases, the interfaces are intimate with no porosity or oxides present. This is because WC and W2C are readily wetted by molten Ni and NiCr. During EAS spraying, the wetting occurred as soon as the Ni and NiCr sheaths began to melt at the wire tips, protecting the Ni–WC/W2C interface from further oxidation. But at the same time, W dissolution can occur easier at this interface. The exposed Ni-rich surfaces however were in contact with oxygen at high temperature, resulting in surface oxide layers around the particles, which became incorporated into the coatings, see Figs. 1 and 2. The amounts of oxide in both coatings show large variations from one location to another with most regions indicating the value of less than 0.1% making quantitative measurement not reliable. Small cracks occurred throughout both coatings. These cracks can be categorized as intergranular and lateral cracks, with the intergranular crack being much more frequent, see Fig. 3. The short lateral cracks can occasionally be observed. The cracks are caused by thermal stress arising from rapid solidification and thermal expansion mismatch between WC/W2C and Ni-rich phases. This is followed by stress relaxation by cracking during cooling from the spraying temperature [15]. Previous works have shown that the presence of Cr in Ni and its alloys has an effect on
Intergranular crack
Lateral crack
Fig. 3. Cracks in WC/W2C in sample b.
their thermal expansion coefficients. Its effect however is small when compared to that of the W [16,17]. W dissolves into Ni-rich phases during the EAS coating process aided by the readily wetting of Ni. The amount of dissolved W depends on the existing amount of Cr in the Ni phase. When there is no Cr, more W can diffuse into Ni. But as Cr increases, a lower amount of W can dissolve into solid solution with Ni–Cr. This is in agreement with the Cr–Ni–W equilibrium phase diagram. At 1200 1C, the solubility limit of W in Ni is as much as 40 wt%.
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
197
Table 3 Thermal expansion coefficients of Ni alloys with various chemical compositions, calculated from a model proposed by Dosovitskiy et al. (2009), and of WC [17,19,20]. Chemical composition (wt%)
Thermal expansion coefficient ( 10 6 1C 1) at 20–800 1C
Ni
Cr
W
WC
Bal. Bal. Bal. Bal. Bal. – –
– 20 – – 20.0 – –
– – 5.4 20.0 5.0 100 –
– – – – – – 100
Intergranular crack Detaching carbide
15.9 15.4 14.6 12.6 14.7 4.5–5.0 4.8
When 20 wt% of Cr is in solution with Ni, however, the W solubility limit in NiCr reduces to lower than 18 wt% [18]. Therefore, there was more W diffused into coating b matrix, making the major phase Ni–W, than there was W diffused into coating a matrix. As much as 20 wt% of W can often be seen in the Ni-rich phase in coating b, occasionally reaching as high as 50%. While only 5 wt% of W is commonly observed in NiCr-rich phase in coating a. W is a refractory metal with the lowest thermal expansion coefficient (TEC) amongst pure metal of about 4.5 10 6 1C 1 at room temperature. When W is added to other metal alloys, it can alter their TECs. The thermal expansion of NiW, NiCr, and NiCrW alloys has been measured by quartz dilatometry for the 20–800 1C temperature range by Dosovitskiy et al. [17]. It was found that substitution of Ni by W leads to a considerable decrease in TEC, while Cr has only a small influence on the TEC of the alloy. The prediction model from this work was utilized in creating Table 3 for the purpose of comparing the TECs of the Ni-rich splats to that of the WC particles [19,20]. In sample b, where it is common to find as much as 20 wt% or more of W in the Ni splats, which corresponds to a drop of TEC to below 12.6 10 6 1C 1. By contrast, it is infrequent that a NiCr splat containing more than 5 wt% of W is detected in sample a. A Ni-rich splat containing 20 wt% Cr and 5 wt% W is corresponding to the TEC of approximately 14.7 10 6 1C 1. Hence, upon rapid cooling from spraying temperature, coating a (NiCrW) suffered higher thermal stress than coating b (NiW) as a result of the higher contraction as well as the thermal expansion mismatch with the WC/W2C particles. The higher thermal stress leads to the weakening of the inter-splat bonds, thus allowing easier propagation of intergranular cracks. The image analysis technique used in measuring the porosity level of the coatings detects both pores and cracks as porosity. The lateral cracks are often very small and so are not picked up as porosity. The intergranular cracks due to the thermal contraction of the coatings, on the other hand, are large enough to be included in the total values of the porosity. Apart from the cracks, there were also small globular pores as a result of incomplete contact between successive splats [13] and voids due to the particle detachment, see Fig. 4. Table 4 shows the levels of porosity in the coatings. The value distributions are large due to the heterogeneous nature of the coatings. The higher percentage of porosity in coating a is a result of a larger amount of the intergranular crack. Furthermore, the stress from cutting and polishing of the sample cross sections, together with the weaker inter-splat bonding due to large thermal stress obtained during coating fabrication in coating a, inevitably dislodged some particles, leaving behind characteristic rounded craters, which were detected as pores, see Fig. 5. In coating b, the thermal stress from coating fabrication also caused the intergranular cracks. The lower thermal expansion mismatch between
Intergranular crack
Carbide
Fig. 4. Intergranular cracks in (a) sample a and (b) sample b.
Table 4 Percentage porosities and hardness values of coatings.
Sample a Sample b
% Porosity
WC/W2C content (vol%)
Microhardness (Hv0.4)
Brinell hardness (Hb)
2.02 7 0.70 0.90 7 0.50
24.37 2.8 17.3 71.9
5727 162 381 7 167
1657 6.4 1517 2.7
Crater left by carbide Fig. 5. Detachment of carbide particle in sample a.
splats can result in lower stress at the intergranular interface and lower carbide detachment, see Fig. 6. Hence, the total porosity is lower in sample b than in sample a.
198
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
Intergranular crack
Crater from carbide detachment
Deformed carbide
Fig. 6. Debonding of carbide particle in sample b. Fig. 8. Sliding wear track of sample a. Table 5 Coating compositions. (2) Calculated %WC (vol%), (3) Measured %WC (1) Coating (vol%), using Image if all W in composition analysis technique (converted to vol% (1) is in WC form from wt%) measured using EDS Ni Sample a 58 Sample b 78.7
Cr
W
19.3 –
22.7 21.3
27.4 26.1
24.3 17.2
Results calculated using density of Ni¼ 8.9 g/cm3, density of Cr¼ 7.1 g/cm3, density of W ¼ 19.25 g/cm3 and density of WC ¼ 15.8 g/cm3.
Crater from carbide detachment
Fig. 9. Sliding wear track of sample b.
with the starting cored wire materials, it is clear that a large percentage of the carbide was loss during spraying. It is not unusual in the thermal spraying process to lose certain phases more than others, particularly if the variation in size and density are large as in this case. In column (3), however, when an image analysis technique was used to determine the volume percent of the carbide phase, it was found that a large decline occurs in sample b. This finding supports earlier results which proposed that the carbide phase decomposed and W dissolved into the Ni-rich phase and that a much larger extent of dissolution occurred in sample b than in sample a. Fig. 7. Wear depths and coefficient of frictions determined by dry sliding wear test.
3.3. Coating hardness 3.2. WC contents WC/W2C content varies from one area of the coating to another due to the heterogeneous structure of the EAS coating. Table 4 shows the WC/W2C contents of the coatings. The average values indicate that, by percentage volume, there is a larger amount of carbides in sample a than there is in sample b. Given that the starting cored wire contains 50% by weight or about 35–36% by volume of WC/W2C, the experimental data of 24.3 and 17.3 vol% for sample a and b, respectively, may appear on a low side. Quantitative EDS was employed to measure the total coating compositions at 600 magnification with the light elements (in this case B, C and O) omitted. The results are presented in Table 5 column (1). In column (2), it was then assumed that all W from column (1) is in the form of WC. In which case, C was manually added to the calculation, resulting in the WC volume percentage of 27.4 and 26.1 for coating a and b, respectively. When compared
Table 4 reports the average hardness values of the coatings. The dissolution of W results in large variation of the hardness throughout the coating. This can be observed when Vickers microhardness is performed. The results show a very large distribution. Indentation size effect (ISE) can be an issue in the utilization of the microhardness test. In this test however the indent size of greater than 10 mm width is comparable to the thickness of the splat rendering ISE less dominant [21]. A better representative of the hardness can be obtained using the macrohardness Brinell method on the surface of the coating. However the indent size in this case is large in comparison to the coating thickness. Hence the hardness of the underlying substrate affects the measurement. Nevertheless Brinell testing takes into account the defects in the coating and can give a better representative of the effective hardness than the Vickers test. The lower Brinell hardness in coating b is due to a larger amount of W being dissolved in the Ni matrix in coating b than there is in the matrix
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
Weight Change (g)
Time (min.) 20
0 -0.002
0
40
60
80
10
20
-0.004 -0.006
sample a
-0.008
sample b
-0.01
Fig. 10. Result from abrasive wear test.
Crater
199
0.15 70.01 mm and 0.12 70.01 mm respectively. The low distributions in the test results indicate the good repeatability of the test. The reason for the difference in the wear depth is related to the intergranular cracks and the detachment of WC/W2C particles. In sample a, where carbide detachment can occur more easily, even though the coating started off with higher amount of carbide particles the stress induced by the sliding action caused the intergranular cracks to propagate and WC/W2C to be pulled-out, allowing the softer matrix to wear more rapidly. Figs. 8 and 9 reveal the wear tracks on the two coatings. The sliding direction is from left to right in both pictures. Both samples show little deformation on the wear tracks. There are, however, a large number of craters on the surface of sample a. These craters are the result of WC/W2C detachments during the sliding wear testing. There are fewer surface craters present on the surface of sample b due to better bonding between splats as a result of lower thermal stress during fabrication. 3.5. Abrasive wear
Carbide
Fig. 11. Abrasive wear surface of sample a.
Ni-W solid solution
Carbide Crater
Fig. 12. Abrasive wear surface of sample b.
Table 6 Average measurement of WC/W2C pull-out after the abrasive wear test. WC/W2C pull-out (vol%) Sample a Sample b
1.72 71.04 2.66 71.03
The results from the three-body abrasive wear testing in Fig. 10 show that sample b again performed better than sample a. The wear surfaces in Figs. 11 and 12 however demonstrate that the reason for this outcome is markedly different from the sliding wear test. In the abrasive case, there are some WC/W2C detachments in both coatings, see Table 6. The extent of the carbide detachment is significantly lower in sample a, when comparing the surfaces after the abrasive test to those after the sliding test. The difference in the carbide pull-out is not as extreme in sample b. This is because the cutting action of the loose alumina abrasives put less strain on the coatings, rendering the weak carbide-matrix bonding not as important as in the sliding case, where deformation and seizure between the ball and the coatings promoted the intergranular crack propagation. If the WC/W2C detachment is minimal then sample a, which contains a larger proportion of carbide and possesses a higher hardness, should outperform sample b in the abrasive wear test. With this contradiction, the focus was turned to the Ni-rich matrix. Cr was originally added to Ni in the cored wire used to produce coating a with an intention that the presence of Cr would promote the formation of hard phases such as Cr3C2 and Cr7C3 which can improve the wear resistance of the coating. However, the short spraying process does not allow for the Cr carbide formation and there was no significant amount of Cr carbide detected in this study. Thus Cr does not significantly contribute to the coating hardness via the formation of Cr carbides. Tungsten as an alloying element in Ni on the other hand has been shown in previous works to have a positive influence on the hardness and fracture toughness, thus improving the abrasive resistance of the material [22,23]. The wear surface of sample b in Fig. 12 shows that the Ni–W phase possesses superior abrasion resistance to the Ni phase, presumably as a result of solid solution strengthening by W, causing the wear surface to be uneven. This Ni–W is thus believed to play a major role in providing the abrasive wear resistance for the coating.
4. Conclusions in coating a. This inevitably results in a greater loss of WC/W2C phase in coating b, thus resulting in the lower hardness. 3.4. Dry sliding wear
For NiCrBSi–WC and NiBSi–WC coatings produced from cored wires using the EAS technique, the results can be summarized as follows:
Sliding wear tests were performed using steel balls on the polished surfaces of the coatings. The results in Fig. 7 show that sample a exhibits higher wear depth than sample b, at
1. W from the WC/W2C dissolved into the Ni-rich matrix during spraying. There was more W content in NiBSi (sample b matrix) than there was in NiCrBSi (sample a matrix), resulting in a
200
P. Sheppard, H. Koiprasert / Wear 317 (2014) 194–200
lower thermal expansion mismatch in NiBSi, hence lower thermal stress from the coating fabrication. The thermal stress in the coatings may propagate intergranular cracks, resulting in more WC/W2C being debonded in the NiCrBSi–WC coating (sample a) where less W was detected in the Ni-rich matrix. 2. The NiBSi–WC coating contained a lower amount of WC/W2C phase than the NiCrBSi–WC coating due to extensive dissolution of W into the Ni-rich matrix. 3. Sliding action during the dry sliding wear test caused more WC/W2C to be detached from the NiCrBSi–WC coating, resulting in higher wear depth than NiBSi–WC coating. 4. Abrasive wear testing showed that the NiBSi–WC coating had slightly higher abrasive resistance than the NiCrBSi–WC coating, even though it contained a lower percentage of carbide than the latter. This is due to the presence of the extensive Ni–W phase.
References [1] S.P. Lu, O.Y. Kwon, Y. Guo, Wear behavior of brazed WC/NiCrBSi(Co) composite coatings, Wear 254 (5–6) (2003) 421–428. [2] S. Stewart, R. Ahmed, T. Itsukaichi, Contact fatigue failure evaluation of post-treated WC–NiCrBSi functionally graded thermal spray coatings, Wear 257 (9–10) (2004) 962–983. [3] A. Martín, J. Rodríguez, J.E. Fernández, R. Vijande, Sliding wear behaviour of plasma sprayed WC–NiCrBSi coatings at different temperatures, Wear 251 (1–2) (2001) 1017–1022. [4] N.Y. Sari, M. Yilmaz, Improvement of wear resistance of wire drawing rolls with Cr–Ni–B–Siþ WC thermal spraying powders, Surf. Coat. Technol. 202 (13) (2008) 3136–3141. [5] L.C. Betancourt-Dougherty, R.W. Smith, Effects of load and sliding speed on the wear behaviour of plasma sprayed TiC–NiCrBSi coatings, Wear 217 (1) (1998) 147–154. [6] J. Przybylowicz, J. Kusinski, Structure of laser cladded tungsten carbide composite coatings, J. Mater. Process Technol. 109 (1–2) (2001) 154–160.
[7] J. He, J.M. Schoenung, A review on nanostructured WC–Co coatings, Surf. Coat. Technol. 157 (1) (2002) 72–79. [8] T.Y. Cho, J.H. Yoon, K.S. Kim, K.O. Song, Y.K. Joo, W. Fang, S.H. Zhang, S.J. Youn, H.G. Chun, S.Y. Hwang, A study on HVOF coatings of micron and nano WC–Co powders, Surf. Coat. Technol. 202 (22–23) (2008) 5556–5559. [9] S.D. Cramer, B.S. Covino Jr. (Eds.), ASM Handbook Vol. 13A, Corrosion: Fundamentals, Testing, and Protection, ASM International, Ohio, 804. [10] P. Wu, C.Z. Zhou, X.N. Tang, Microstructural characterization and wear behavior of laser cladded nickel-based and tungsten carbide composite coatings, Surf. Coat. Technol. 166 (2003) 84–88. [11] K.V. Acker, D. Vanhoyweghen, R. Persoons, J. Vangrunderbeek, Influence of tungsten carbide particle size and distribution on the wear resistance of laser clad WC/Ni coatings, Wear 258 (2005) 194–202. [12] P. Niranatlumpong, H. Koiprasert, Phase transformation of NiCrBSi–WC and NiBSi–WC arc sprayed coatings, Surf. Coat. Technol. 206 (2011) 440–445. [13] S. Deshpande, A. Kulkarni, S. Sampath, H. Herman, Application of image analysis for characterization of porosity in thermal spray coatings and correlation with small angle neutron scattering, Surf. Coat. Technol. 187 (1) (2004) 6–16. [14] Standard test method for linearly reciprocating ball-on-flat sliding wear, G 133-95, Annual Book of ASTM Standards, 03.02, ASTM, 1995, pp. 529–536. [15] H. Herman, Plasma sprayed coatings, Sci. Am. 259 (3) (1988) 112–117. [16] F. Xiao, R. Yang, L. Fang, C. Zhang, Solidification shrinkage of Ni–Cr alloys, Mater. Sci. Eng. B 132 (2006) 193–196. [17] G.A. Dosovitskiy, S.V. Samoilenkov, A.R. Kaul, D.P. Rodionov, Thermal expansion of Ni–W, Ni–Cr, and Ni–Cr–W alloys between room temperature and 800 1C, Int. J. Thermophys. 30 (2009) 1931–1937. [18] K.P. Gupta, The Cr–Ni–W system, Phase Diagrams of Ternary Nickel Alloys, Monograph Series on Alloy Phase Diagrams, The Indian Institute of Metals, Calcutta (1990) 111–133. [19] H. Doi, Y. Fujiwara, K. Miyake, Y. Oosawa, A systematic investigation of elastic moduli of WC–Co alloys, Metall. Trans. 1 (5) (1970) 1417–1425. [20] E. Lassner, W.D. Schubert, Tungsten: Properties, Chemistry, Technology of the Element, Alloys and Chemical Compounds, Kluwer Academic, Plenum Publishers, New York (1999) 31. [21] I. Manika, J. Maniks, Size effects in micro- and nanoscale indentation, Acta Mater. 54 (2006) 2049–2056. [22] C.A. Schuh, T.G. Nieh, H. Iwasaki, The effect of solid solution W additions on the mechanical properties of nanocrystalline Ni, Acta Mater. 51 (2) (2003) 431–443. [23] K.T. Kim, J.H. Lim, J.H. Kim, J. Joo, W. Nah, B.K. Ji, B.H. Jun, C.J. Kim, G.W. Hong, Development of cube-textured Ni–W alloy substrates for YBCO-coated conductor, Phys. C: Supercond. 412–414 (2) (2004) 859–863.