Tribological behaviour of enamel coatings

Wear 426–427 (2019) 319–329

Contents lists available at ScienceDirect

Wear journal homepage: www.elsevier.com/locate/wear

Tribological behaviour of enamel coatings

T

⁎

Huynh H. Nguyen, Shanhong Wan , Kiet A. Tieu, Sang T. Pham, Hongtao Zhu School of Mechanical, Materials, Mechatronic and Biomedical Engineering, University of Wollongong, Northfield Avenue, Wollongong, NSW 2522, Australia

A R T I C LE I N FO

A B S T R A C T

Keywords: Enamel coating Wear Friction Sliding wear

An experimental study has been carried out to evaluate the tribological and mechanical properties of steel hardened with glass-ceramic enamel under ambient conditions. Three types of borosilicate based enamels were fused onto mild steel by the firing process, after a sandblasting and a ground coat to ensure surface integrity between the substrate and the covering coat. The surface roughness, micro hardness, wear track and tribological characteristics of the enamel coating were then characterised by an atomic force microscopy, a Vicker microhardness tester, a 3D optical profiler, an EDS/SEM, and a friction and wear testing tribology meter. The tribological differentials of these top coats were examined after sliding against chrome steel, zirconia and silicon nitride. The worn morphologies provided some understandings of the failing mechanism due to friction. During dry sliding with a silicon nitride mating part, the coefficient of friction and loss of enamel enriched with titanium was less than the chrome steel balls, although the zirconia ball succumbed to serious wear. The changes in the worn morphologies and chemical composition under different sliding contacts, and the corresponding wear mechanism were observed, with abrasive wear being dominant.

1. Introduction Vitreous or porcelain enamel has been in use as far back as the Egyptian period. Enamel is fundamentally a combination of a metallic substrate and a glassy coating that is achieved by melting and fusing glass to a metal substrate in a temperature range between 720 °C and 870 °C [1], or even higher; in this range the thermal curing creates a vitreous enamel with a glossy finish that is heat resistant up to 1000 °C [2]. This heat treatment process during manufacturing can crystallise some compositions within the coatings [3–6], such that the enamel can be either glassy (like glass) or like glass-ceramic; the latter has superior properties due to a combination of glass and ceramic phases. An enamel coating has outstanding properties such as extreme hardness [5–7], high temperature and thermal shock resistance [8,9], chemical inertness [10], anti-corrosion, anti-oxidation [11–18], and resistance to abrasion and scratching [19–25]. Thanks to these promising thermal, physico-chemical and mechanical properties, enamel has been used in applications as various as usage in daily life, as well as industrial and engineering applications such as food and chemical processing, automobiles, petrochemicals, thermal power plants and aeronautical engineering [26]. Steel hardened with enamel resists abrasions and scratches, characteristics that enable the surface to withstand mechanical impact; this ability is known as composition-dependent. Many studies dedicated to

⁎

the relationship between property, structure, and composition have enabled researchers to understand the anti-abrasion and anti-scratch properties of enamel coatings. For example, according to Rossi et al. [21], mill additions such as potassium feldspar and zirconium silicate undermine the abrasion resistance of enamel due to the increased roughness and dissolubility of mill agents within the modified enamel structure. However, the additions of spodumene, feldspar, zirconium silicate and quartz promotes the abrasion resistance of enamel because they improve densification if the amount added does not exceed a threshold [22] and the elimination of porosity is advantageous to abrasion resistance [27,28]. On the other hand, the large size and dissolubility of these additives do lead to a deterioration in abrasion [22]. Although enamel coatings are hard and resistant to abrasion, their low fracture toughness affects their abrasive wear behaviour because it is associated with brittleness induced fracture [20]. During tribological testing, enamel can protect metallic substrates from damage caused by severe mechanical impact [29–31]. According to Zhang et al. [29], enamel coatings can improve the wear characteristics by 3.12 times more than the substrate and by an additional 1.64 times when rare earth oxides are added; this is due to the high hardness of enamel and the porosity free inner structure stemming from the adjusted composition. Moreover, enamel also enhances the already significant wear resistance due to the four fold drop in mass between the samples with and without a coating; this occurs because the hardness and the

Corresponding author. E-mail address: [email protected] (S. Wan).

https://doi.org/10.1016/j.wear.2019.02.002 Received 3 September 2018; Received in revised form 28 January 2019; Accepted 1 February 2019 0043-1648/ © 2019 Elsevier B.V. All rights reserved.

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

roughness reduce the apparent contact area and damage caused by fretting [31–33]. This high roughness also reduces the contact area between the sliding counterparts which also reduces the friction [30]. The common conclusion from these studies on the wear mechanism of enamel coating is abrasive wear. There are so many studies in literature which are dedicated to investigating the abrasion resistance of enamel coatings, but only a few on their tribological behaviour, and even they are limited to the tribology of a type of borosilicate sliding against chrome steel. It is therefore important to understand the tribological characteristics of different enamel materials against different counterparts, so in this study, three types of enamels base on commercial borosilicate and three different balls are used as counterparts to observe their tribological behaviour. These enamels differ in terms of their chemical compositions, with one being rich in titania and the others having zirconium and cobalt, respectively. Their counterparts include chrome steel, zirconia and silicon nitride balls. The tribological behaviour of these enamel-ball tribo pairs were tested under drying sliding conditions and then the results were compared; this study will provide the knowledge needed to manufacture enamel coatings for engineering applications such as the mining industry.

Table 2 Properties of commercial balls used as stationary tribo part [34]. Property 3

Density (d, g/cm ) Young's modulus (E, GPa) Hardness (H, GPa) Surface Roughness (Ra, nm)

GCr15 (GC)

ZrO2 (Z)

Si3N4 (Si)

7.8 270 7 20

6.0 205 14 25

3.44 320 15 10

investigated mechanically via micro scratch testing with a progressive mode loading from 0.01 to 100 N for 1 min, while the resulting 3 mm scratches are observed by SEM. An atomic force microscope (AFM), with a scanning size of 50 × 50 µm2 and a scanning rate of 2 Hz is used to measure the roughness of the coating. The phase compositions of the enamel coating are identified by X-ray diffraction (XRD) with a Cu Kα X-ray source at a step size 0.02° and a scan rate of 1°/min, with a maximum high voltage of 35 kV and a current of 28.4 mA. The hydrophilicity of the enamel coating is determined by the sessile drop method using 2 μL of distilled water. The contact angle is measured with a goniometer attached to an optical camera on at least three positions on the coating surface and then the average values are reported.

2. Materials and methods 3. Results The coatings used in this study are made from borosilicate based enamels supplied from W.G. Ball Ltd., United Kingdom. The enameling materials have been fused at approximately 830 °C for 3.5 min onto mild steel samples with the dimension of 116 × 30 × 1.5 mm3, those have been sandblasted prior to enameling. The coating is slid against two types of ceramic balls (Si3N4 and ZrO2) and a chrome steel ball (GCr15) which are stationary counterparts; these balls are 6.35 mm in diameter. Table 1 and Table 2 show the chemical compositions of three different enamels used for covering coats (denoted as W, B, and Y) and the properties of the commercial countering balls (noted as G, Zr and Si). Ball-on-plate tests are carried out at room temperature on a Bruker UMT TriboLab tribometer under dry reciprocating conditions; these tests last for 900 s at an applied load of 10 N which is equivalent to the maximum Hertzian contact pressure of approximately 0.73 GPa, a rotatory speed of 4.8 Hz, and a 27 mm long stroke. The normal and lateral forces are automatically recorded by coupled sensors and then transferred to a computer system which calculates the coefficient of friction for the whole travel distance of 233.28 m in 900 s. The choice of the test time 900 s is to obtain the long-term trend of friction. Three tests took place in order to collect a representative of the friction-time curve and an average coefficient of friction for each tribo pair. The worn surfaces of the counterparts are observed using a JEOL 6490 scanning electron microscope (SEM) attached to an energy dispersive spectroscope (EDS). The accelerating voltage is 15 keV and the working distance is fixed at 10 mm; a Bruker ContourGT-K 3D Optical Microscope is also used to inspect the morphology of the worn surfaces and to determine the wear volume. Before commencing sliding testing, the hardness of the as-prepared coatings is measured using a Vickers microhardness tester with an applied load of 0.49 N in 10 s dwell time. These measurements are taken at 9 points on each sample and then the average value becomes the accepted hardness of the coating. The enamel coating is also

3.1. Characterisation of the enamel coating The AFM results show that the as-prepared enamel coatings are smooth, with a roughness of approximate 16–30 nm (Ra), albeit their surfaces are slightly wavy, and the Rz values are quite low and between 180 and 310 nm. The as-prepared enamel coatings are hydrophilic with very low contact angles of 16–17°. These values are similar to the results given by Chen et al. [35]. Fig. 1 shows that the water droplets spread readily onto the as-obtained surfaces. The hardness of enamel coatings is shown in Table 3; Y is the hardest of the three types of enamels, followed by B and W, respectively, but they are still rather brittle. Fig. 2 shows that the acoustic emission (AE) of all the enamel coatings increased in the early stage but without any scratches on the surface, so the damage under the top layer occurred due to a possible bubble structure within the coating (Fig. 4), a brittle coating, and low fracture toughness [36]. The scratch becomes visible under a load of 14 N when it would then be easy to expand laterally. The scratches of enamel Y and B are similar in width, and larger than that of W. The saturated AE shows complete destruction which confirms the brittle fracture failure of these tested enamel coatings. Fig. 3 shows the XRD patterns on different enamels after thermal fusion; the pattern of coating B has glassy characteristics and a gentle hill without peaks, whereas the W and Y enamels have a baseline similar to B but with many diffraction peaks that represent the crystalline phases. These features are attributed to anatase titanium dioxide [37] and zirconium silicate [38] in samples W and Y, respectively. The results of XRD agree with the SEM images obtained from the top surfaces of samples etched with hydrochloric acid. The SEM images depict the presence of many small particles and some larger particles on the surfaces of coatings W and Y (Fig. 4d & e), whereas there is no special feature on coating B (Fig. 4f).

Table 1 EDS analysis of three types of borosilicate based cover coats. wt%

SiO2

B2O3

TiO2

Na2O

K2O

CaO

ZnO

Al2O3

CoO

ZrO2

P 2 O5

W Y B

40.39 50.00 51.22

14.35 10.72 14.25

18.15 6.37 4.57

7.58 17.23 16.82

7.36 1.22 1.43

– 2.00 4.39

0.73 – –

9.37 7.54 7.25

– – 0.07

– 4.92 –

1.73 – –

320

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al. 30

a)

25

16

17

o

o

Titania

16

Substrate

20

Contact angle ( o)

d)

Ground coat

o

Cover coat 15

10

b)

5

e) Zircon

0 W

B

Y

Fig. 1. Contact angle of as-produced enamel coatings. Table 3 Hardness of as-produced enamel coatings.

c)

Enamel

W

Y

B

Hardness (H, GPa)

5.23 ± 0.48

6.92 ± 0.29

6.33 ± 0.26

a)

a)

Fig. 4. Optical microscopy images of cross-sections (a, b, c), and SEM images of enamels etched with hydrochloric acid (d, e, f) in samples W, Y and B, respectively. b)

AE (%)

The presence of crystals is believed to enhance the properties of the glassy base thanks to the combined characteristics of crystal and glass [5,6], properties that are dependent on different crystalline compositions. In fact, the Titania has a Mohs’ hardness of 5.5, while that of Zircon is 6.5–7.5. Zircon (zirconium silicate) improved the hardness of enamel Y better than Titania did to sample W, which is why the hardness of coating Y shown in Table 3 is very high. The optical microscope images (Fig. 4a–c) show the layers of enamel layers have a bubbled structure, and there are many large bubbles within the ground coat of every enamel. These features emerge near the substrate, then move and increase in size towards the top coat, while some of them penetrate into the top layer. Most of the bubbles are less than 15 µm, although some reach 20–40 µm, and several reach 50 or 80 µm; there are also bubbles in the cover coat, but they are smaller than those in the intermediate coat. While the ground coat helps to bond the substrate and top coat, it usually has a low viscosity to wet the substrates. A ground coat with low viscosity is needed to constrain any large air voids. Whereas coating W has the least number of bubbles inside the layer compared to the others, coatings Y and B do have bubbles but they are finer than in coating Y because the titanium acts as a fluxing agent, zirconia is in the refractory, and cobalt increases the melt viscosity. In accordance with the optical microscope images above, the two layer coating varies from 400 to 500 µm thick, while the average values for each single layer of enamel are 186 ± 13, 221 ± 2 and 162 ± 6 µm for coatings W, Y and B, respectively. The ground coating is 252 ± 17 µm thick.

c)

0

20000

14 N

40000

60000

80000

100000

Fn (mN)

Fig. 2. Micro-scratch image of different enamels: B (a), Y (b), and W (c). Scale bar is 500 µm. AE range is 0–100%.

3.2. Friction and wear Figs. 5 and 6 show that when the two counterparts began to make contact during the sliding test the coefficient of friction of each sample was at its highest in the early running-in stage, after which it decreased

Fig. 3. XRD pattern of enamel (a) W, (b) Y and (c) B.

321

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

0

(a)

-50

(b)

Wear depth (μ m)

-100 -150 -200 -250

(a) Si3N4_W (b) GCr15_W (c) ZrO2_W

-300

(c) -350 0

500

1000

1500

2000

2500

3000

3500

Wear width (μm) Fig. 5. Coefficient of Friction of different balls sliding against enamel W.

0

(b)

Wear depth (μ m)

-10 -20

(a)

-30 -40

(a) Si3N4_W (b) Si3N4_Y (c) Si3N4_B

-50

(c) -60 400

800

1200

1600

2000

Wear width (μm) Fig. 8. Cross-sectional histogram of the wear track of coatings.

tribological behaviour with the enamel samples in the reciprocating test; after a sharp increase induced by the running-in stage, the coefficient of friction rapidly decreased in the first 65 m, but then all the enamels remained around 0.55 in the later stage. The friction signal of GCr15 was of a steady decline whereas the zirconia friction was steady for 120 m and then dropped towards at the end of 900 s experiment. Such occurrence can be observed in the repeated tests. There is a possibility of different topographical behaviors on the samples if they are allowed to run longer distances. This will be considered in the future work. Moreover, the friction lines in Fig. 5 fluctuate more for the cases of zirconia and GCr15. This fluctuation is caused by the wear mechanism that depends on the interfacial interaction between zirconia and GCr15 and different surface topography as shown in Fig. 9. Fig. 5 shows the coefficient of friction as the chrome steel ball slid against coating W was higher than silicon nitride by 0.15 whereas Fig. 6 shows that the coefficient of friction with the tribo-pairs of various enamel coatings sliding against a silicon nitride ball did not differ very much; in reality the chrome steel surface produces a result that is similar to other studies [29,31], but the silicon nitride counterparts have a better wear performance. Fig. 7 shows that the zirconia ball led to severe wear in enamel W of 13.93 mm3, which is approximately 7.8 times more than by the GCr15 ball. This agrees with the results where the coefficient of friction in terms of the ZrO2-W tribo pair was smaller at the later stage of the test (Section 4). Meanwhile, the wear loss of coatings W and Y against silicon nitride were similar at 0.37 and 0.33 mm3, respectively, which is one third as much as enamel B at 1.09 mm3, even though the respective coefficients of friction of 0.3, 0.56, and 0.55 did not differ very much. Ball wear is also shown in Fig. 7. Note that the degree of wear

15

20.0

14

17.5

13

15.0

12

12.5

11

10.0

1.5

7.5

1.0

5.0

0.5

2.5

Wear loss of ball (10-3mm3)

Wear loss of coating (mm3)

Fig. 6. Coefficient of Friction of different enamels counter-reciprocating against a silicon nitride ball.

0.0

0.0 GCr15_W

ZrO2_W

Si3N4_W

Si3N4_Y

Si3N4_B

Pair of counterparts Fig. 7. Wear loss of coating and ball after reciprocating test.

at different rates depending on the material of the balls. GCr15 gave a gradual decrease in friction but had the coefficient of friction of 0.69, and while the friction of zirconia was lower and stable at 0.60 during the first 130 m, it then decreased and stabilised at 0.45 for the remainder of the test; the average coefficient of friction in this case was 0.53. The silicon nitride ball displayed an interesting type of

322

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

a)

b)

c)

e)

f)

a) d)

bubbles b) g)

h)

i)

k)

l)

n)

o)

c) j)

d) m)

e) Delamination

Grooves

Fig. 9. 3D Profilometer and SEM images (left to right) of wear track on the enamel coating: a-c) GCr15-W, d-f) ZrO2-W, g-i) Si3N4-W, j-l) Si3N4-Y and m-o) Si3N4-B.

3.3. Wear track morphologies of coatings

between the silicon nitride-enamel sliding pairs was opposite to the coatings, but there was very little difference in the degree of wear. Coating Y had the least wear, while the counterpart experienced the highest wear at 11.4 × 10−3 mm3. In contrast, coating B experienced the most wear, with the wear on its counter ball being 9.5 × 10−3 mm3. The corresponding loss as the silicon nitride ball slid against coating W was 10.7 × 10−3 mm3. The pair of zirconia balls and enamel W experienced the most severe damage on both sliding parts, with the wear on the ball being 13.2 × 10−3 mm3. Although the wear of coating W against the GCr15 ball was higher than coating W, the silicon nitride pair, the corresponding wear on the steel chrome ball was almost half that of the silicon nitride. The cross-sectional profile of the wear track shown in Fig. 8 indicates an agreement with the results of the loss through wear and the morphology (Section 3.3). The lines are fluctuating due to the presence of crystals and bubbles within the coating and delaminated parts on the wear tracks.

Fig. 9 shows the morphologies of wear tracks of coatings after tribological testing where the surfaces of all the coatings were covered with exposed bubbles. In terms of the tribo pairs with enamel W (Fig. 9(a-i)), as coating W slid against the zirconia ball it experienced the most severe damage and the widest wear track (Fig. 9(d-f)), but when sample W slid against silicon nitride it produced the narrowest wear track (Fig. 9(g-i)); the remaining tribo pairs with chrome steel balls also showed severe wear but in the middle range (Fig. 9(g-i)), unlike the other two pairs above. Fig. 9(a-c) shows the wear track has a rough surface and many parallel grooves, which is similar to the chrome steel ball (Fig. 10a). Meanwhile, enamel W sliding against the zirconia counterpart shows an inner bubble structure among the spalls and pits on the worn surface, whereas those sliding against silicon nitride balls have smooth wear tracks. The wear track morphologies of different types of enamel coating in the tribo pairs with the same silicon nitride balls have similar features, 323

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

a)

b)

c)

d)

e)

f)

g)

h)

i)

j)

k)

l)

m)

n)

o)

Built up material

Cavities

Grooves

Trapped materials

Peeling off

Fig. 10. SEM and 3D profilometer images (left to right) of balls used in sliding pairs: a-c) GCr15-W, d-f) ZrO2-W, g-i) Si3N4-W, j-l) Si3N4-Y and m-o) Si3N4-B.

coating, therefore the greater amount of delaminated material on enamel B is also due to the low fracture toughness, or extreme brittleness of the glass coating. The smaller amount of delamination on enamel Y is due to the presence of a few zircon crystals, whereas the titania-rich phase helps W resist delamination better than the two above. The grooves on the worn surface are due to the hard debris from the counterparts, while the large ploughed tracks on the enamel coating tested with a silicon nitride surface are attributed to the large amount of delaminated debris from the coating. As a result, enamel coatings suffer from abrasive wear due to their brittleness, as well as from delaminated parts and roughened and spalled surfaces as shown in Fig. 9o. In fact these worn surfaces also have “ploughed” tracks caused by debris as the third body; this outcome agrees with Fan et al. [39]. Such occurrence of wear is associated with scuffing happening in the first 100 s (or approximately 26 m), this scuffing will be considered in future work.

including a few ploughed tracks and evidence of delamination. However, the delaminated parts of enamel B are the largest, and whereas the wear track of enamel Y is generally smoother than that of W, coating Y has delaminated all through the wear surface. The delaminated spalls on coating W appear to be less than on coating Y, and while a bubble structure can facilitate delamination because of the width of the wear on W and Y, it is similar to but narrower than the wear on coating B. Obviously, the width of the wear track worn on the surfaces of enamel corresponds to the width of the horizontal wear scar on their ball counterparts, so the width of the wear track on the coatings will be the same as the width of the ball scar (Section 3.4). Some features seen on the ball such as grooves and roughness are also present on the enamel coatings because all the wear tracks on enamel coatings that slid against silicon nitride balls are due to the polishing effect of hard silicon nitride debris as a third body. While the roughness of W in the Wchrome steel tribo pair is due to titania crystals inside the coating, the delamination of worn enamel surfaces can be induced by the brittle 324

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

Fig. 11. EDS mapping image of partial wear scar on GCr15 ball sliding against coating W. The sliding direction is along the vertical axis.

direction. Since the wear scars on balls are elliptical they differ from each other depending on the materials of balls and coatings in the tribo pairs. Although zirconia is harder than chrome steel, it has the largest area of wear, because it is both wider and longer in vertical and horizontal directions, respectively. The SEM (Fig. 10d, e) and EDS mapping images (Fig. 12) mean that this large wear scar on the zirconia ball is probably due to the effect of wear-induced debris from the enamel coating that act as third bodies in the contact. Zirconia as a ceramic oxide might actively interact with the oxide debris of enamel under reciprocating conditions to form strongly bonded compacts with high shear bond strength [40] that can attach itself onto the surface of the ball and roughen the wear scar. These compacted layers like a “strong adhesive” during testing that will be peeled off with the top layer of zirconia, as shown in Fig. 10e, to expand the boundary of the wear-damaged area. Figs. 10, 11, 13, 14 and 15 show that the chrome steel and silicon nitride balls have also got the attached material on their surfaces; however, these two balls do not behave like zirconia that actively reacts with such material, they are inert enough to allow the enamel debris to become partially attached to the chrome steel ball, or trapped on the silicon nitride ball into the cavities produced while the ball is being manufactured. These cavities inside a silicon nitride ball will expand as they are filled with wear debris and by the sliding test conditions. The cross-sectional images of these tested balls are shown in Fig. 16; the balls made from chrome steel and zirconia reveal a dense phase, while the silicon nitride ball has a porous phase.

3.4. The wear scar morphologies of balls As Fig. 10 shows, the wear scars of all the used balls are oval, and the width of wear perpendicular to the sliding direction decreased in the order of zirconia, chrome steel, and silicon nitride balls. The wear scar of the silicon nitride ball sliding against coating B was wider than with coatings Y and W. The wear scar of GCr15 ball had many smooth grooves those were partially covered with some built-up materials (Fig. 10(a-c)), and also showed plastic deformation along the grooves. However, the wear scar of the ZrO2 ball was rougher and appeared to peel off while being tested, as shown in Fig. 10(d-f). Fig. 10(g-o) shows similar wear scars on the Si3N4 parts sliding against different enamel coatings; these scars included smooth grooves and cavities that were either empty or full of debris. The EDS maps (Figs. 11–15) present the elemental distribution on part of the ball wear scar; they also show the built-up materials along the grooves on the chrome steel ball, on the zirconia surface and in the cavities of the silicon nitride ball. This material is debris from the enamel coatings and contains the elemental constituents of coatings such as alkali, silicon, titanium, zirconium and cobalt. The wear scars on the balls are not circular due to the intrinsic brittleness of enamel materials leads to the fractures on the enamel coatings emerging and growing laterally under a loaded contact. The images of micro-scratches on the enamels shown in Fig. 2 support this deduction so the apparent contact area between two sliding counterparts during reciprocating sliding is expected to increase in a lateral 325

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

Fig. 12. EDS mapping image of partial wear scar on ZrO2 ball sliding against coating W.

Fig. 13. EDS mapping image of partial wear scar on Si3N4 ball sliding against coating W.

mentioned previously, i.e., zirconia, chrome steel, and silicon nitride balls. This analysis reveals that the wear mechanism is probably a combination of adhesive and abrasive for chrome steel and zirconia balls, and mostly abrasive wear for silicon nitride. The GCr15 ball experienced abrasive wear due to the continuous parallel grooves caused by the debris and anatase phases, together with plastic deformation along the grooves followed by the adhesive wearing action of these deformed parts. Meanwhile, the zirconia ball appeared to undergo plastic deformation before either adhesive or abrasive expansion as the edge of the wear scar peeled off. The silicon nitride ball experienced a combination of abrasion and adhesion because it had been ploughed by the debris to form grooves, while the materials trapped in the cavities

With regards to the wear scars on the silicon nitride ball, they increased in width when the balls slid against the W, Y and B enamels, respectively (Fig. 10c–e). This increase in the order of wear area is due to the brittleness induced fracture of the enamel coatings. Although coatings Y and B are much harder than W, they have a low fracture toughness, as shown qualitatively in Fig. 2 (Section 3.1). The scratches on coatings Y and B are wider than on coating W, possibly because the large number of titania crystals were impeding the lateral expansion of the scratched tracks. Although crystalline phases are also present within enamel Y, there are not enough to constrain the scratched track compared with the enamel W. Since the lateral growth of fractures on enamels W, Y and B is different, there will be an increase in the contact width, or the width of the wear scar of the balls in the same order 326

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

Fig. 14. EDS mapping image of partial wear scar on Si3N4 ball sliding against coating Y.

Fig. 15. EDS mapping image of partial wear scar on Si3N4 ball sliding against coating B.

4. Discussion

inside the ball also act as third bodies to abrade and expand the void during the test.

The highest loss of enamel on the W zirconia ball can be attributed to the superior hardness of the ball and the highly compacted layer of enamel on its surface. In general, the contact areas between the zirconia 327

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

friction. The friction in the chrome steel case gradually decreased as the grooves became smoother during the testing period, which led to a reduction in the contact area. Similarly, the zirconia ball also exhibited high friction with coating W, although it was less than the chrome steel ball. This occurred because the zirconia is harder than the steel so the crystal phases within enamel W could not roughen up the surface of the ball. Furthermore, due to the formation of a strongly bonded compact layer of enamel debris on the surface of the ball the zirconia counterpart caused severe wear on enamel W and also made a macro rough wear track on the coating which probably caused the contact area to fluctuate in size and quickly reduce the coefficient of friction. The silicon nitride ball had a cavity structure that increased the contact area between the ball and titania particles with the enamel coating once sliding commenced, so the friction in the early stage was rather high, even higher than the zirconia ball. However, the debris filled these defective regions and reduced the contact area and led to a sharp fall in friction. Since silicon nitride was much harder than the other materials in this study, it produced a polishing effect that facilitated smoothening the worn coating and thereby a lower coefficient of friction. The trend in the other tribo pairs of silicon nitride ball with different enamel coatings was similar, so there was no distinct difference in friction by enamels W, Y and B, and therefore the small deviations can be attributed to the mechanical capability of enamel and the variations of chemical compositions within the enamel coating.

Fig. 16. Optical microscopy images of the cross-section of different balls.

surface and the enamel coating undergo local heating due to the adiabatic character of zirconia [41]; this causes the zirconia to undergo plastic deformation and the distinct possibility of surface veneering with the ceramic debris from the enamel coating. The resulting products are strongly bonded layers [40] with a very rough surfaces that cause severe abrasive wear on the coating, as shown in Fig. 9b. The zirconia counterpart also experiences very severe damage as the compacted layer becomes detached and the material on the balls is peeled off. This process stems from a combined failure of the compacted layers of veneered zirconia and veneering ceramic debris during sliding, as well as a cohesive spallation in the built-up, or veneering, materials, and adhesive delamination at the ceramic-zirconia interfaces [40,42,43]. Meanwhile, the wear on enamel W when sliding against chrome steel is greater than with the silicon nitride balls, whereas their counterparts experienced an opposite trend. It can be deduced that the enamel debris from metallic oxides (e.g. iron, chrome) from chrome steel and the coating are not as hard as the ball, so the surface of ball is not worn as much by the debris. In contrast, ceramic debris from silicon nitride can be as hard as the ball, so as the third bodies wear out there is more debris from the silicon nitride ball than the chrome steel ball. Enamel W is protected from wear by the hard silicon nitride-related debris which reduces the amount of wear on the coating. Moreover, coating W experienced more wear from the chrome steel ball, as evidenced by the wider wear track, while coating W paired with chrome steel ball experienced much more wear than the silicon nitride surface. In terms of different enamel coatings sliding against silicon nitride balls, sample B experienced the most damage due to brittleness, or low fracture toughness, and as a consequence, the longest and widest wear track (Section 3.1). However, the wear on the counter ball was the smallest because, (1) enamel Y was harder and wore out the ball of silicon nitride more than coating B, and (2) although coating W was not hard and area of wear was narrower than coating B, the large amount of crystalline titania particles increased the roughness of the wear track, and thus the rough worn surface and the silicon nitride debris caused more severe wear on the ball due to the combined roughness and hard third bodies. In contrast, the wear of coatings W and Y were similar and rather lower than the coating B because the crystalline phases enhanced the fracture toughness of the coatings. A small difference in wear between W and Y and their counterparts can be attributed to the superior hardness of coating Y. With regards to the friction, the chrome steel ball produced the highest coefficient among the tribo pairs of enamel W with three different ball materials. This was due to the large number of anatase titania particles which, together with the debris, ploughed and produced a continuous groove pattern on the surface of the ball. This increased the rough wear surfaces on the enamel coating and the ball, which also increased the apparent contact area, the result was a high coefficient of

5. Conclusions The capability assessments and tribo-surface analysis of steel specimens coated with enamel lead to the following conclusions: – Enamel coatings maintain relatively high friction characteristics, depending on the counterparts. The enamel coating with a high concentration of titania delivered less friction when sliding against silicon nitride whereas zirconia in contact with steel surfaces hardened with enamel had higher friction. – Enamel coatings suffer mainly from abrasive wear with ploughed grooves, distinct delamination, and spallation. Chrome steel and zirconia surfaces undergo a combined adhesive and abrasive action, whereas the silicon nitride ball experienced mostly abrasive wear. The Zirconia surface interacted actively with the enamel materials to form a compacted layer that caused severe wear and higher friction on both sliding counterparts. – The existence of crystalline phases inside enamel coatings can enhance their mechanical strength. The scratch comparison reveals that the enamel coating enriched with titania had the desired fracture toughness, and protected the coatings from severe wear. Zircon phases also provided wear resistance to enamel coating by improving the hardness of the coating, however the enamel with pure glass phase caused severe abrasive wear due to its high brittleness and lower fracture toughness. Acknowledgments The first author is grateful for the financial support from the Engineering Mechanics Centre at UOW through a scholarship. The authors acknowledge financial support from the Australian Research Council for the two projects DP170103173 and LP160101871. The authors thankfully acknowledge the assistance from technical staffs in Electron Microscopy Center (EMC) facilities, University of Wollongong. References [1] S.P. Rodtsevich, S.Y. Eliseev, V.V. Tavgen, Low-melting chemical resistant enamel for steel kitchenware, Glass Ceram. 60 (2003) 25–27. [2] K.A. Maskall, D. White, Vitreous Enamelling: A Guide to Modern Enamelling

328

Wear 426–427 (2019) 319–329

H.H. Nguyen, et al.

Practice, Published on behalf of the Institute of Ceramics by Pergamon Press, 1986. [3] A. Goel, E.R. Shaaban, F.C.L. Melo, M.J. Ribeiro, J.M.F. Ferreira, Non-isothermal crystallization kinetic studies on MgO-Al2O3-SiO2-TiO2 glass, J. Non-Cryst. Solids 353 (2007) 2383–2391. [4] H. Kotan, K.A. Darling, M. Saber, C.C. Koch, R.O. Scattergood, Effect of zirconium on grain growth and mechanical properties of a ball-milled nanocrystalline FeNi alloy, J. Alloy. Compd. 551 (2013) 621–629. [5] D. Wang, Effect of crystallization on the property of hard enamel coating on steel substrate, Appl. Surf. Sci. 255 (2009) 4640–4645. [6] A. Bachara, A. Mabrouk, D.D.S. Meneses, E. Veron, Y. Sadallahd, P. Echeguta, F. Bentisse, Effect of thermal treatment on the property of enamel coating on steel substrate, Mater. Environ. Sci. 8 (2017) 3884–3891. [7] B.T. Garland, The designing of products utilising the special properties of vitreous enamel, Mater. Des. 7 (1986) 44–48. [8] A.K. Varshneya, Chapter 12 - Thermal Conductivity and Heat Transfer in Glass, in: A.K. Varshneya (Ed.), Fundamentals of Inorganic Glasses, Academic Press, San Diego, 1994, pp. 233–240. [9] A.K. Varshneya, Chapter 11 - Heat Capacity of Glass, in: A.K. Varshneya (Ed.), Fundamentals of Inorganic Glasses, Academic Press, San Diego, 1994, pp. 225–232. [10] A.K. Varshneya, Chapter 17 - Chemical Durability, in: A.K. Varshneya (Ed.), Fundamentals of Inorganic Glasses, Academic Press, San Diego, 1994, pp. 397–408. [11] F. Tang, G. Chen, R.K. Brow, J.S. Volz, M.L. Koenigstein, Corrosion resistance and mechanism of steel rebar coated with three types of enamel, Corros. Sci. 59 (2012) 157–168. [12] F. Tang, G. Chen, J.S. Volz, R.K. Brow, M. Koenigstein, Microstructure and corrosion resistance of enamel coatings applied to smooth reinforcing steel, Constr. Build. Mater. 35 (2012) 376–384. [13] F. Tang, X. Cheng, G. Chen, R.K. Brow, J.S. Volz, M.L. Koenigstein, Electrochemical behavior of enamel-coated carbon steel in simulated concrete pore water solution with various chloride concentrations, Electrochim. Acta 92 (2013) 36–46. [14] F. Tang, G. Chen, J.S. Volz, R.K. Brow, M.L. Koenigstein, Cement-modified enamel coating for enhanced corrosion resistance of steel reinforcing bars, Cem. Concr. Compos. 35 (2013) 171–180. [15] F. Tang, G. Chen, R.K. Brow, Chloride-induced corrosion mechanism and rate of enamel- and epoxy-coated deformed steel bars embedded in mortar, Cem. Concr. Res. 82 (2016) 58–73. [16] D. Zheng, S. Zhu, F. Wang, Oxidation and hot corrosion behavior of a novel enamelAl2O3 composite coating on K38G superalloy, Surf. Coat. Technol. 200 (2006) 5931–5936. [17] Y.M. Xiong, S. Zhu, F.H. Wang, Effect of ultrafine enamel coating on the oxidation and mechanical property of Ti60 alloy, Acta Metall. Sin. 40 (2004) 768–772. [18] L. Xin, F. Wang, T. Li, S. Zhu, Y. Xiong, Effect of enamel coating on long-term oxidation of Ti65 alloy, Int. J. Mater. Product. Technol. 20 (2004) 327–334. [19] P.F.A. Testing, resistance of enameled surfaces to scratching gouging, and abrasion, J. Am. Ceram. Soc. 30 (1947) 94–104. [20] I.M. Hutchings, Tribology: Friction and Wear of Engineering Materials, Edward Arnold, 1992. [21] S. Rossi, E. Scrinzi, Evaluation of the abrasion resistance of enamel coatings, Chem. Eng. Process. 68 (2013) 74–80. [22] S. Rossi, C. Zanella, R. Sommerhuber, Influence of mill additives on vitreous enamel properties, Mater. Des. 55 (2014) 880–887. [23] S. Rossi, N. Parziani, C. Zanella, Abrasion resistance of vitreous enamel coatings in function of frit composition and particles presence, Wear 332–333 (2015) 702–709. [24] S. Rossi, M. Fedel, F. Deflorian, N. Parziani, Abrasion and chemical resistance of

[25] [26] [27]

[28] [29]

[30] [31]

[32]

[33]

[34]

[35]

[36]

[37] [38]

[39]

[40]

[41] [42] [43]

329

composite vitreous enamel coatings with hard particles, Surf. Interface Anal. 48 (2016) 827–837. E. Scrinzi, S. Rossi, The aesthetic and functional properties of enamel coatings on steel, Mater. Des. 31 (2010) 4138–4146. A. Majumdar, S. Jana, Glass and glass-ceramic coatings, versatile materials for industrial and engineering applications, Bull. Mater. Sci. 24 (2001) 69–77. K. McKinley, H. Evele, C. Baldwin, Analysis of fracture in porcelain enamels, in: Proceedings of the 22nd International Enamellers Congress, IEI, Cologne - Germany, 2012. R.W. Cahn, R.J. Brook, Processing of Ceramics, VCH, 1996. S. Zhang, Y.H. Ren, M.R. Sun, F. Hu, C.H. Zhang, Effects of RE on the friction and abrasion character of Porcelain Enamel, Coat., Adv. Mater. Res. 538–541 (2012) 410–413. Z. Pala, N. Ganev, K. Kolařík, O. Bláhová, J. Jersák, Surface integrity and tribological behaviour of hardened steels, Chem. Listy 105 (2011) s739–s742. Q. Feng, Y. Zhang, Z. Dong, P. Zhang, J. Wang, D. Wang, Q. Gao, X. Peng, F. Wang, Preparation and Wear Behavior of Enamel Coating on Ti-6Al-4V ELI Titanium Alloy, in: Y. Han (Ed.), Advances in Materials Processing, Springer, Singapore, Singapore, 2018, pp. 47–59. Y. Fu, N.L. Loh, A.W. Batchelor, D. Liu, Z. Xiaodong, J. He, K. Xu, Improvement in fretting wear and fatigue resistance of Ti-6Al-4V by application of several surface treatments and coatings, Surf. Coat. Technol. 106 (1998) 193–197. Y. Fu, H. Du, Effects of the counterface materials on the tribological characteristics of CNX coating deposited on plasma-nitrided Ti-6Al-4V, Mater. Sci. Eng.: A 298 (2001) 16–25. S. Wan, B.H. Tran, A.K. Tieu, Y. Xia, H. Zhu, S. Cui, Q. Zhu, The influence of water addition on high-temperature tribological properties of interstitial free steel sliding against different counterparts, Tribol. Trans. 61 (2018) 713–725. L.Y. Chen, Q.Q. Fang, W.Z. Jiang, X.P. Liao, J.Y. Ye, H.L. Jin, Study of the hydrophobic properties of the surface of an enamel material, Mater. Sci. Forum 789 (2014) 178–182. K. Hrabovska, J. Podjuklova, O. Životský, K. Barcova, I. Štěpánek, V. Bartek, T. Lanik, Effect of indication liquids on brittle-fracture properties of vitreous enamel coating, Chem. Listy 105 (2011) s743–s747. L. Aldon, P. Kubiak, A. Picard, J.C. Jumas, J. Olivier-Fourcade, Size particle effects on lithium insertion into Sn-doped TiO2 anatase, Chem. Mat. 18 (2006) 1401–1406. J. Liu, J. Song, T. Qi, C. Zhang, J. Qu, Controlling the formation of Na2ZrSiO5 in alkali fusion process for zirconium oxychloride production, Adv. Powder Technol. 27 (2016) 1–8. L. Fan, H.H. Chen, W.T. Mao, J.J. Cao, Friction and wear behavior of CaO-MgOAl2O3-SiO2 glass-ceramics prepared by waste casting sands, Appl. Mech. Mater. 217–219 (2012) 1251–1256. M. Sehgal, A. Bhargava, S. Gupta, P. Gupta, Shear bond strengths between three different yttria-stabilized zirconia dental materials and veneering ceramic and their susceptibility to autoclave induced low-temperature degradation, Int. J. Biomater. 2016 (2016) 9658689. M.-S. Suh, Y.-H. Chae, S.-S. Kim, Friction and wear behavior of structural ceramics sliding against zirconia, Wear 264 (2008) 800–806. İ. Tuncel, P. Özat, E. Eroğlu, Effects of coloring procedures on zirconia/veneer ceramics bond strength, J. Adv. Prosthodont. 6 (2014) 451–455. B.-K. Choi, J.-S. Han, J.-H. Yang, J.-B. Lee, S.-H. Kim, Shear bond strength of veneering porcelain to zirconia and metal cores, J. Adv. Prosthodont. 1 (2009) 129–135.