Two-body abrasive wear of nano- and microcrystalline TiC–Ni-based thermal spray coatings

www.elsevier.com/locate/surfcoat

Surface & Coatings Technolog

Two-body abrasive wear of nano- and microcrystalline TiC–Ni-based

thermal spray coatings

X. Qia,*, N. Eigena, E. Austa, F. Gartnerb, T. Klassena, R. Bormanna

aGKSS Research Centre, Max-Planck-Str. 1, 21502 Geesthacht, GermanybUniversity of the Federal Armed Forces, Holstenhofweg 85, 22043 Hamburg, Germany

Received 17 January 2005; accepted in revised form 13 May 2005

Available online 6 July 2005

Abstract

In the present study, the wear behaviour of nanocrystalline coatings of the composition (Ti,Mo)(C,N)–45 vol. % NiCo, prepared by

vacuum plasma spraying (VPS) and high-velocity oxy-fuel (HVOF) spraying of high-energy-milled powder, is characterized and compared

to microcrystalline coatings of the same composition. Two-body abrasive wear tests, as well as scratch tests, are applied to produce wear

traces on the surfaces of the nano- and microcrystalline coatings. While nanocrystalline HVOF coatings are weaker than their

microcrystalline counterparts, nanocrystalline VPS coatings show superior wear resistance. The worn surface morphologies are investigated

with optical microscopy, scanning electron microscopy and atomic force microscopy. The wear mechanisms and failure of nano- and

microcrystalline coatings are distinctly different and are discussed in detail.

D 2005 Elsevier B.V. All rights reserved.

Keywords: Wear mechanism; Nano- and microcrystalline coatings; TiC–Ni; Thermal spray; Two-body abrasive wear

1. Introduction

Wear-protective coatings are mostly based on WC–Co

and Cr2C3–NiCr composites, with hard phase particles of

some micrometers embedded in a metallic matrix [1]. While

WC–Co composites lead to a high wear resistance and can

be used only at moderate temperatures [2–4], Cr2C3–NiCr

composites can guarantee high hot gas corrosion resistance.

However, compared to WC–Co, the wear resistance of

Cr2C3–NiCr is about an order of magnitude lower [5–8].

TiC–Ni-based cermets can bridge the gap between WC–Co

and Cr2C3–NiCr with comparable good performance under

abrasive and corrosive attack [1,4,9,10]. In addition, TiC as

the hard phase has a lower density and is less expensive

[11]. The TiC–Ni-based composites can be alloyed with

other elements, tailoring the composition for the needs of

different processes and applications [12]. Alloying Co to the

0257-8972/$ - see front matter D 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.surfcoat.2005.05.007

* Corresponding author. Tel.: +49 4152 87 1944; fax: +49 4152 87 1927.

E-mail address: [email protected] (X. Qi).

binder phase increases the hardness of the matrix. Adding

Mo to the hard phases or metallic alloys enhances the

adhesion between ceramics and binder and positively

influences oxidation kinetics. Moreover, from sintered

ceramics, it is known that replacing C by N of up to 20%

in the hard phase limits carbide growth during sintering and

leads to a more uniform phase distribution [12,13]. Unlike

the WC–Co composites, TiC shows significantly higher

thermal and thermodynamic stabilities and can lead to less

undesired reaction with metallic matrixes and less decom-

position of nanocrystalline carbide phases. The respective

TiC-based composites should be ideal to study size effect

without the disturbance of side reactions.

In order to improve the wear resistance of cermet

materials, a combination of high toughness and high

hardness is desired [14,15]. Such a combination has been

achieved by the reduction of the hard phase particle sizes in

sintered WC–Co cermets [16,17]. Furthermore, nanocrys-

talline WC–Co composite material with very high hardness

possesses a higher wear resistance compared to conven-

tional material [18–22]. Therefore, it is expected that the

y 200 (2006) 5037 – 5047

X. Qi et al. / Surface & Coatings Technology 200 (2006) 5037–50475038

wear resistance of thermal spray composite coatings can be

also improved by using nanocrystalline cermet precursor

material. However, in contrast to bulk material, coatings

may show processing flaws, which influence the wear

resistance. For instance, Usmani et al. [23] and Stewart et al.

[24] have reported poorer sliding and abrasive wear of high-

velocity oxy-fuel (HVOF) sprayed nanocrystalline WC–Co

coatings compared to microcrystalline coatings under a

variety of wear test conditions. Qiao et al. [25,26] produced

coatings of WC–Co showing inferior or superior wear

resistance of nanocrystalline coatings compared to conven-

tional coatings, depending on the spray parameters used.

Kear et al. [19,20] and Zhu et al. [27] reported that low-

pressure plasma-sprayed nanocrystalline WC–Co compo-

site coatings display superior hardness and wear resistance

compared to microcrystalline coatings.

Recently, coatings of the composition (Ti,Mo)(C,N)–45

vol.% (Ni–20 wt.% Co) were produced by a combination of

high-energy milling and thermal spraying [28]. It was

shown that the nanocrystalline feedstock material produced

by high-energy milling was not significantly coarsened

during thermal spraying. In the coating, the crystallite size

of most hard phases and matrix were in the range of 10–100

nm. Although it was likely that a fraction of carbide phase

might be dissolved in the binder during the spray process,

the result showed that in the deposited nanocrystalline

coating, only very minor amount of carbide were dissolved.

Therefore, the hard phases could be mainly maintained in

the material [28].

In order to highlight specific advantages of these nano-

crystalline coatings, in the present paper, nano- and micro-

crystalline (Ti,Mo)(C,N)–45 vol.% (Ni–20 vol.% Co)

HVOF and vacuum plasma spraying (VPS) coatings are

investigated, which are based on high-energy milled or

agglomerated and sintered composite powders. Two-body

abrasive wear tests and scratch tests are carried out to evaluate

the wear resistance of the coatings. The wear mechanisms are

discussed on the basis of microstructure, hardness, wear

morphologies of the coatings, as well as the morphology of

the wear debris and the mass loss during wear tests.

2. Experimental

Nanocrystalline (Ti,Mo)(C,N)–45 vol.% (Ni–20 vol.%

Co) powders were prepared by high-energy ball milling.

Commercially available (Ti,Mo)(C,N), Ni and Co powders

with a mean particle size of FSSS 1.5–2.5 Am were mixed

and processed in a horizontal attrition mill, a modified

simoloyer CM08 (Zoz GmbH, Wenden, Germany) under

argon, using hardened bearing steel balls as milling media.

The details of the milling processes and the properties of the

powders were described in previous publications [28–30]. It

was shown in Refs. [28,29] that the carbide phase and the

metallic matrix in the high-energy milled powders had the

crystallite sizes of about 10 nm. Microcrystalline powder

was prepared by agglomerating and sintering of the same

initial material as used for the high-energy milled nano-

crystalline powders. Subsequently, nano- and microstruc-

tured (Ti,Mo)(C,N)–45 vol.% NiCo cermet powders were

used as feedstock for thermal spraying. HVOF spraying was

performed with a Sulzer Metco Diamond Jet 2700 torch by

using oxygen and ethylene in a slightly under-stoichiometric

ratio to obtain maximum temperatures and velocities of the

gas jet. For the VPS process, a modified Medicoat Mach 3

system was used with an Ar–He–H2–plasma gas mixture

to guarantee optimum velocities. The coatings were sprayed

onto mild steel substrates. The exact spraying parameters

are given in previous publications [31–33].

Abrasive wear tests were carried out according to the

Japanese standard JIS H8615 [30]. In this standard test, a

metallic wheel, wrapped with a 320-grit SiC abrasive paper,

is pressed against the coating with a load of 30 N and is

moved back and forth over a distance of 30 mm. After each

double stroke (DS), the wheel is rotated by 0.9-. The

abrasive wear is determined by the mass loss of the

specimen (in mg) after 1200 DS. In addition to the 320-

grit SiC paper (main SiC particle size 45 Am), variant JIS

H8615 tests using 500 grit (main SiC particle size 30 Am),

150 grit (main SiC particle size 100 Am) and 80 grit (main

SiC particle size 200 Am) were carried out to investigate the

influence of the load of individual abrasive particle on wear

behaviours of the nano- and microcrystalline coatings.

Furthermore, for a deeper understanding of the interaction

between the individual SiC abrasive particles and the

coating and to eliminate the influence of wear debris, single

grooves were prepared by applying only one DS on the

polished coating surfaces. In addition, single grooves were

also prepared on the polished coating surfaces by scratching

using a standardized Vickers indenter. These scratch tests

were performed with a Micro Scratch Tester (MST) at CSM

Instruments SA, Peseux, Switzerland. Different loads and a

scratch speed of 3 mm/min were applied.

Microhardness measurements were conducted on pol-

ished cross sections of the coatings using a Vickers indenter

and a load of 100 g (HV 0.1) according to DIN EN ISO

6507. The values presented are an average of 30 readings.

For the microstructural analysis, surfaces and polished

cross sections of the coatings were investigated by optical

microscopy (OM) using an Olympus PMG 3, Olympus

Corporation, Tokyo, Japan, and scanning electron micro-

scopy (SEM) using a Zeiss DSM 962, Carl Zeiss AG,

Oberkochen, Germany, in back-scattered electron (BSE)

mode. Detailed information about the surface morphology

was obtained by SEM in secondary electron (SE) mode.

Additionally, the morphologies of the worn coating surfaces

and the profiles of wear grooves after different abrasive

wear tests were recorded by atomic force microscopy

(AFM). An AFM of type Dimension 3000, Digital Instru-

ments, Inc., Santa Barbara, USA, was employed using a tip

of type TAP300/RTESP, NanoDevice, Inc, Santa Barbara,

USA, with a tip radius <10 nm. The measurements were

X. Qi et al. / Surface & Coatings Technology 200 (2006) 5037–5047 5039

conducted in tapping mode in order to keep the mechanical

influence of the tip on the surface morphology negligible.

With the tapping mode, the AFM measures the morphology

by slightly tapping, instead of contacting, the surface with

the tip.

3. Results

3.1. Coating microstructure

Fig. 1 shows optical micrographs of VPS and HVOF

spray coatings in cross section on the basis of nanocrystal-

line and microcrystalline powders. All coatings have a

thickness of about 200 Am and bond well to the substrate.

The nanocrystalline VPS coating (Fig. 1a) shows a very

homogeneous microstructure and a porosity of less than 1%.

The different phases in the coating cannot be distinguished

in this magnification. In the microcrystalline VPS coating

(Fig. 1b), the hard phase particles (dark grey) are large

enough to be resolved. Some pores (black), typically larger

than the hard phase particles, are visible in the coating.

More quantitative investigations by image analysis reveal

that the porosity of the microcrystalline coating amounts to

only about 6% which was significantly higher than that of

the nanocrystalline coating (1%) [34]. The microstructure in

the nanocrystalline HVOF coating (Fig. 1c) is less

homogeneous than in the nanocrystalline VPS coating and

contains thin dark layers inhomogeneously distributed

between the splats. In the microcrystalline HVOF coating

(Fig. 1d), the dark layers seem to be finer and more

homogeneously distributed. Previous investigations [31,34]

by X-ray diffraction and energy-dispersive spectroscopy

(EDS) have shown that the dark layers and black spots

mainly contained TiO2.

ba

c d

50 µm

50 µm

Fig. 1. OM micrographs of the nano- and microcrystalline VPS- and HVOF-coat

VPS-coating, (c) nanocrystalline HVOF-coating, and (d) microcrystalline HVOF-

For the interpretation of the wear behaviour, the initial

microstructure of the coating has to be characterized first.

Respective micrographs of the polished surfaces are shown

in Fig. 2. The nanocrystalline VPS coating (Fig. 2a) shows

pancake-like splats with only little porosity at the splat

boundaries. The hard phase particles are very fine and

homogeneously distributed within the binder matrix. Pre-

vious investigations have shown that the typical size of the

hard phase particles in the nanocrystalline VPS coating is

about 100 nm [34]. In the microcrystalline VPS coating

(Fig. 2b), hard phase particles in the range of 2–5 Am can be

observed and the porosity is higher compared to that of the

nanocrystalline coating. In the nanocrystalline HVOF coat-

ings, oxide layers are thicker and predominantly located

around the splats (Fig. 2c), whereas in the microcrystalline

HVOF coating, the oxides are more homogeneously

distributed, and thus, no individual splats can be distin-

guished (Fig. 2d). Moreover, in the HVOF coatings less

(Ti,Mo)(C,N) hard phase particles are visible due to the loss

of C by oxidation.

3.2. Coating hardness

The Vickers hardnesses (HV 0.1) of VPS and HVOF

coatings sprayed with nano- and microcrystalline feedstock

powders are compared in Fig. 3. Independent of the spray

processes, nanocrystalline coatings exhibit higher hardness

than microcrystalline coatings. Compared to the nano-

crystalline VPS coatings, the nanocrystalline HVOF coat-

ings are softer due to the partial oxidation of hard phase

particles and respective loss of C, thus reducing the total

amount of the hard phase remaining in the coating. On the

contrary, the microcrystalline HVOF coating is slightly

harder than the microcrystalline VPS coating, probably due

to the higher porosity of the latter.

50 µm

50 µm

ings in cross section: (a) nanocrystalline VPS-coating, (b) microcrystalline

coating.

a b

dc

10 µm 10 µm

10 µm 10 µm

Fig. 2. SEM micrographs (BSE mode) of nano- and microcrystalline VPS- and HVOF-coatings in top view: (a) nanocrystalline VPS-coating, (b)

microcrystalline VPS-coating, (c) nanocrystalline HVOF-coating, and (d) microcrystalline HVOF-coating. The BSE contrast reveals the metallic binder bright,

carbides grey, oxides dark and pores black.


3.3. Wear resistance according to JIS H8615 tests

The mass loss after two-body abrasive wear tests (JIS

H8615) of nano- and the microcrystalline coatings obtained

by VPS and HVOF spraying is compared in Fig. 4. The

difference in the wear rates between the VPS and HVOF

coatings is remarkable. Prepared by using the same feed-

stock powder, the HVOF coatings show the mass loss of

about 3–4 times that of the VPS coatings. Despite the

higher hardness of the nanocrystalline coatings, they show

similar or even higher mass loss than the respective

microcrystalline coatings for VPS and HVOF processing.

Fig. 5 shows wear surfaces of the nano- and the

microcrystalline VPS and HVOF coatings after the JIS

H8615 tests. Continuous grooves can be observed in the

nanocrystalline VPS coating (Fig. 5a), and the grooves

600

750

900

1050

1200

VPS-coating HVOF-coating

Har

dn

ess

[kg

mm

-2]

nanocrystalline coatingmicrocrystalline coating

Fig. 3. Hardness (HV0.1) of the nano- and microcrystalline (Ti,Mo)(C,N)–

45 vol.% NiCo coatings sprayed by VPS- and HVOF-processes. The

presented values are an average of 30 readings.

generally appear shallower and more homogeneously

distributed than those in the microcrystalline VPS coating

(Fig. 5b). The surface roughness (Ra) of the nanocrystalline

VPS coating is measured to be only about one-half that of

the microcrystalline VPS coating after the JIS H8615 tests.

From the worn surface of the microcrystalline coating, two

different types of morphology can be distinguished:

(1) Zones of huge, continuous grooves (an example is

indicated by the circle 1 in Fig. 5b);

(2) Zones of small, discontinuous grooves (an example is

indicated by the circle 2 in Fig. 5b), separated by large

hard phase particles.

Pores that were present in the unworn VPS coating

cannot be detected after wear testing. They have obviously

0

10

20

30

40

50

60

HVOF-coatingVPS-coating

nanoscrystalline coatingmicrocrystalline coating

Mas

s lo

ss [

mg

/120

0DS

]

Fig. 4. Mass loss of the nano- and microcrystalline (Ti,Mo)(C,N)–45 vol.%

NiCo VPS- and HVOF-coatings after the abrasive wear test (JIS H8615).

c d

a b

20 µm 20 µm

20 µm 20 µm

1

2

Fig. 5. SEM micrographs (SE mode) of the worn surfaces of the nano- and microcrystalline VPS- and HVOF-coatings after the JIS H8615 tests: (a)

nanocrystalline VPS-coating, (b) microcrystalline VPS-coating, (c) nanocrystalline HVOF-coating, and (d) microcrystalline HVOF-coating.


been closed due to the plastic deformation of the coating

material during wear. In the HVOF coatings (Fig. 5c and d),

besides grooves, also some defects with irregular shape exist

(examples are indicated by the arrows in Fig. 5c and d).

Apparently, the defects on the surface are material delami-

nations caused by the applied stress from the abrasive

particles. In accordance with the higher material loss in wear

tests, material delaminations can only be observed in HVOF

coatings and occur more frequently in the nanocrystalline

HVOF coating (Fig. 5c).

Such material delaminations can also be observed in

cross sections of the HVOF coatings. The example shown in

Fig. 6 was prepared from a nanocrystalline HVOF coating

with the surface that is ground with SiC abrasive papers and

polished sequentially with 3 Am, 1 Am and 0.25 Amdiamond pastes. Two pieces of the coating material, each

in the size and shape of a splat (indicated by the white

arrows in Fig. 6), are separated from the coating along the

splat boundaries. These delaminations may be attributed to

the grinding and polishing procedure during specimen

preparation. This finding confirms that oxide inter-splat

20 µm

Fig. 6. Optical micrograph of the cross section microstructure with

delaminated material (indicated by white arrows) at the polished surface

of a nanocrystalline HVOF-coating.

layers are highly subjected to the mechanical failure in the

HVOF coatings.

3.4. Wear resistance according to variant JIS H8615 tests

In order to attain more detailed information on the

interaction of the abrasive particles with the coating

material, grooves were produced by the JIS H8615 equip-

ment by applying only one double stroke on a polished

nanocrystalline coating surface. SEM micrographs of such

grooves show the formation of continuous grooves in

nanocrystalline coatings (Fig. 7a and c). Comparing Fig.

7a with c, the VPS and HVOF coatings show about the

same amount of grooves, which are arbitrarily distributed

like the abrasive particles on the grinding paper. The

comparison of typical groove morphologies of the nano-

crystalline VPS and HVOF coatings, as shown in Fig. 7b

and d, reveals that the grooves in the VPS coating show

more plastically deformed material piled up at the ridges

than for the groove of the HVOF coating. Additionally, in

the HVOF coating material delaminations are observed, as

already seen on the surface after 1200 DS (an example is

indicated by Fe_ in Fig. 7c). Delaminations are located

predominantly along the grooves. The close-up view (Fig.

7e) shows the lack of material at the groove. Small

imperfections can also be observed in the VPS coating

(Fig. 7a). However, their distribution is random and

independent of wear grooves indicating that they have no

correlation to delamination during the wear tests and can be

attributed to initial pores formed in the spraying process.

Morphologies obtained by AFM of the typical wear

grooves and the corresponding cross section profiles of

100 µm 5 µm

5 µm

c

e

c

e

a

100 µm

b

5 µm

b

d

Fig. 7. SEM (SE mode) micrographs of the morphologies of wear traces produced by the JIS H8615 equipment (abrasive particle size 45 Am, normal load 30 N,

1 DS) on a polished surface of the nanocrystalline VPS- and HVOF-coatings: (a and b) VPS-coating, (c), (d and e) HVOF-coating. The figure in (b), (d) and (e)

show the close-up microstructures at the places marked by Fb_, Fd_ and Fe,_ respectively in (a) and (c).


nanocrystalline coatings are depicted in Fig. 8. The different

widths can be attributed to locally different loads of the

abrasives. Aside from the groove in the VPS coating, ridges

have built up that protrude from the surface level. The total

0 1 2 3 4 5 6-160-120-80-40

04080

VPS-groovedept

h [n

m]

width [µm]

(a)

Fig. 8. AFM analysis of single-groove morphologies on the nanocrystalline VPS- a

the micrographs indicate the places where the profiles are located. The wear traces

load 30 N, 1 DS) on the polished coating surfaces.

section area of both ridges is about half of the section area of

the groove, which means that half of the material was cut

and half was plastically deformed and pushed to both sides

of the groove. In the HVOF coating, ridge formation is

0 1 2 3 4 5-160-120-80-40

04080

HVOF-groovedept

h [n

m]

width [µm]6

(b)

nd HVOF-coatings: (a) VPS-coating, (b) HVOF-coating. The white lines in

were produced by the JIS H8615 tester (abrasive particle size 45 Am, normal

0

10

20

30

40

200 µm100 µm45 µm30 µm

nanocrystalline VPS coatingmicrocrystalline VPS coating

Abrasive particle size

Mas

s lo

ss [

mg

/ 12

00 D

S]

Fig. 10. Mass loss of the nano- and microcrystalline VPS coatings after

abrasive wear tests using the JIS H8615 equipment with different grit


negligible, which means that nearly all material was cut and

removed from the coating as wear debris.

The morphology of the grooves produced by 1 DS on a

polished surface in the microcrystalline VPS coating is

shown in Fig. 9. According to the load and shape of

individual abrasive particles on the grinding paper, grooves

with different sizes are produced in the coating (Fig. 9a).

The close-up view of a small groove shows that the abrasive

particle causes significant material displacement in the

metallic matrix, while it only leaves a very small trace in

big hard phase particles (Fig. 9b). In contrast, in the large

grooves, the hard phase particles are fragmented and the

width of the groove appears homogeneous and independent

of the different phases in the groove (Fig. 9c).

From the above results, the load-dependent groove sizes

might be a decisive factor to influence the wear mechanism

of the microcrystalline coating. Therefore, in order to

8 µm

c

5 µm

2 µm 0.8 µm

b

5 µm

20 µm

a

Fig. 9. SEM micrographs (SE mode) of grooves in microcrystalline VPS

coatings: (a) overview, (b) a small groove, and (c) a large groove close to a

small groove. The grooves were produced by applying 1 DS on the polished

coating surface (JIS H8615 test, abrasive particle size 45 Am, normal load

30 N). The widths of the grooves in (b) and (c) are marked by the numerals

on the micrographs.

papers. Test parameters other than abrasive size are the same as those used

in the standard JIS H8615 test.

investigate the effect of load, another variation of the JIS

H8615 test was carried out by using different grit abrasive

papers. When using a coarse abrasive paper, the applied

1.2 µm0.7 µm

2.7 µm 1.4 µm

5.1 µm

b

a

c

A

3 µm

3 µm

3 µm

Fig. 11. SEM micrographs (BSE mode) of the grooves scratched by the

Vickers indenter at loads of (a) 0.01 N, (b) 0.02 N, and (c) 0.1 N in the

microcrystalline VPS coating. The hard phase particle marked by FA_ in (b)

is partially fractured. The widths of the grooves in the hard and binder

phases in (a) and (b) and in the composite (c) are marked by numerals. The

scratch direction is from left to right.


load is shared by a smaller number of abrasive particles. For

a constant contact area of the coating and the counterbody,

this results in a higher effective load on each abrasive

particle.

Wear test results on the nano- and microcrystalline VPS

coatings using abrasive papers with different grit sizes are

shown in Fig. 10. Mass loss increases with increasing size of

abrasive particles. However, the increasing rate in mass loss

is lower for the nanocrystalline coating than that for the

microcrystalline coating. For instance, when the abrasive

particle size is changed from 30 Am to 200 Am, the mass

loss of the nanocrystalline coating is increased by about

300%, whereas the increase for the microcrystalline coating

is more than 700 %. Therefore, compared to the micro-

crystalline coating, the nanocrystalline coating shows a

superior wear resistance against larger abrasive particles

while it shows an inferior wear resistance against smaller

abrasive particles. Under the present experimental condi-

tions, the critical mean abrasive particle size, above which

the nanocrystalline coating shows higher wear resistance

e f

a b

c d

10 µm

10 µm

20 µm

Fig. 12. SEM micrographs (SE mode) of single-scratch grooves in the nano- and m

(b) microcrystalline VPS-coating, (c) nanocrystalline HVOF-coating, (d) microcry

debris from nanocrystalline HVOF-coating. The grooves were produced by a Vick

direction is from top to bottom.

than the microcrystalline coating, lies between 45 Am and

100 Am.

This result reveals that wear depends on the load carried

by individual abrasive particles. However, the load and

shape of individual abrasive particles cannot be controlled

in the JIS H8615 tests. Therefore, a standardized Vickers

diamond indenter under a defined load is employed for a

quantitative analysis of wear.

3.5. Single scratches produced by using a Vickers indenter

Details concerning the wear mechanism of the micro-

crystalline VPS coating, particularly the fracture of hard

phases, can be revealed by the BSE mode of the SEM. Fig.

11 shows the wear behaviour of the hard phase particles in

the microcrystalline coating scratched at different loads. The

diamond indenter is sharp enough to produce grooves in

large hard phase particles even under the load of 0.01 N.

The indenter can cut away a small fraction of the hard phase

material without breaking the whole particle and leaves a

10 µm

10 µm

20 µm

icrocrystalline VPS- and HVOF-coatings: (a) nanocrystalline VPS-coating,

stalline HVOF-coating, (e) debris from nanocrystalline VPS-coating, and (f)

ers indenter under the load of 1 N on the polished coating surfaces. Scratch


groove in the hard phase particle. However, the width of the

grooves in the hard phase particles is smaller than that in the

binder phase (Fig. 11a). If the load is increased to 0.02 N,

the groove widths in the binder phase and in the hard phase

particles are increased (Fig. 11b). Some of the hard phase

particles are partially fractured (marked by FA_ in Fig. 11b).

Whereas the groove width in the fractured hard phases is not

significantly lowered in comparison with that in the binder,

it is reduced to about a half in the non-broken carbides. If

the load is further increased up to 0.1 N, most of the hard

phase particles in the groove are severely fractured. The

width of the groove appears to be constant and independent

of the phase distribution in the groove (Fig. 11c). It might be

worth noting that small hard phase particles with diameters

less than about 0.5 Am seem not to be fractured under the

applied load.

Fig. 12 shows the SEM micrographs of the grooves

generated by scratching with a Vickers indenter on the VPS

and HVOF coatings at a load of 1 N. The smaller width of

the grooves in the nanocrystalline coatings with respect to

those in the microcrystalline coatings correlates with their

higher hardness. While the groove in the nanocrystalline

VPS coating is very homogeneous (Fig. 12a), there are

cracks in the hard phase particles and at the side of the

groove in the microcrystalline VPS coating (Fig. 12b). It is

evident that the fragmented material at the side of the groove

can be easily removed in subsequent scratches. In the

nanocrystalline HVOF coating, severe material delamina-

tions at the groove are observed (Fig. 12c). Material

delamination, although significantly less, is also found at

the groove in the microcrystalline HVOF coating (Fig. 12d).

The wear debris observed in the grooves of the nanocrystal-

line VPS and HVOF coatings is distinctly different. The

more ductile debris from the VPS coating is long and curved

(Fig. 12e), while the brittle debris from the HVOF coating

has the shape of isometric polygons (Fig. 12f).

-20 -15 -10 -5 0 5 10 15 20-1,5

-1,0

-0,5

0,0

0,5

1,0

coating

surface

B

A

nanocrystalline coatingmicrocrystalline coating

Distance from groove centre [µm]

Dep

th [

µm]

Fig. 13. AFM analysis of the cross section profiles of wear grooves in the

nano- and microcrystalline VPS coatings. The grooves were produced by

the Vickers indenter under the load of 1 N on the polished coating surfaces.

(A) Cross section area of the groove in the nanocrystalline coating; (B)

cross section area of the groove in the microcrystalline coating.

AFM recorded cross section profiles of grooves

scratched by the Vickers indenter at the load of 1 N in the

nano- and the microcrystalline VPS coatings are compared

in Fig. 13. The groove in the nanocrystalline coating is

shallower and narrower than that in the microcrystalline

coating. The cross section area of the groove in the

nanocrystalline coating (area of the zone marked by FA_ inFig. 13) is only about 55% of that in the microcrystalline

coating (area of the zone marked by FB_ in Fig. 13), which

means that the nanocrystalline VPS coating is much more

scratch resistant at 1 N than the microcrystalline VPS

coating.

4. Discussion

4.1. Wear mechanisms of nano- and microcrystalline VPS

coatings

The experimental results obtained from the two-body

wear tests and from the scratch tests on the nano- and

microcrystalline VPS coatings reveal that during the JIS

H8615 tests by using smaller abrasives, the mass loss of

the nanocrystalline coating is higher than that of the

microcrystalline coating (Fig. 10), which means the nano-

crystalline coating is less wear resistant; however, the JIS

H8615 tests by using larger abrasives demonstrate a higher

wear resistance of the nanocrystalline VPS coating (Fig.

10); the scratch tests show that the volume loss in a single

groove in the nanocrystalline coating is much lower than

that of the microcrystalline coating (Fig. 13), which also

indicates a higher wear resistance for the nanocrystalline

VPS coating.

These results can be understood on the background of

different wear mechanisms in the microcrystalline coating

under different test conditions. In the JIS H8615 test by

using smaller abrasives, since the average load carried by

individual abrasive particles is low, most of the hard phase

particles in the microcrystalline coating are not fractured.

The wear resistance of the hard phase particles is higher

than that of the binder phase, and thus, the wear mechanism

is the preferential removal of the soft binder phase. The

loss of the binder phase leads to protrusion of the hard

phase particles. The protruding hard phase particles then

further support the abrasive particles and carry the load,

thus lowering the load of other abrasive particles and,

therefore, the respective penetration depths. This results in

an efficient wear protection using the microcrystalline

coating. However, if the average load carried by individual

abrasive particles is high, e.g., in the JIS H8615 tests by

using larger abrasives and in the scratch tests by a Vickers

indenter at the loads �0.1 N (Fig. 11c), most of the hard

phase particles in the grooves will be fractured and the

fragmented hard phase particles are removed together with

the binder phase. The two mechanisms operate during two-

body wear in different local zones lead to an inhomoge-


neous wear morphologies in the microcrystalline coating

(Fig. 5b).

In contrast, the nanocrystalline coating shows a very

homogeneous wear morphology leading to the conclusion

that only one wear mechanism is operating. Since the hard

phase particles in the nanocrystalline coating are very small,

they cannot support the abrasive particles. Consequently, the

wear mechanism of the nanocrystalline VPS coating is a

combination of cutting and ploughing. The nanoscale hard

phase particles and the binder phase are simultaneously

displaced and removed by abrasive particles like in a

homogeneous rather ductile material, forming wear debris,

in which the hard phase particles are embedded in the binder

matrix. This also explains the slightly lower wear resistance

of the nanocrystalline coating in comparison to micro-

crystalline coatings, as measured with the JIS H8615 tests

by using smaller abrasives, in which a lower average load is

applied by individual abrasive particles as compared to the

scratch test.

However, in the nanocrystalline coating, even under high

local loads, breaking of the hard phase particles is very

unlikely because of the high strength of very fine ceramic

particles [35]. Moreover, being surrounded by a ductile

matrix, the nanoscale hard phase particles are exposed to

smaller local stress concentrations than the larger ones.

Therefore, at higher loads the nanocrystalline coating is

advantageous over microcrystalline coatings; the formation

of areas severely damaged by wear is avoided, thus leading

to a higher wear resistance and a smoother worn surface.

4.2. Wear mechanisms of nano- and microcrystalline HVOF

coatings

The result from the JIS H8615 tests (Fig. 4) shows that

the mass loss of the HVOF coatings is several times higher

than that of VPS coatings processed with the same feedstock

powders and tested under the same wear conditions. This

can be explained by the occurrence of material delamina-

tions during the abrasive wear of the HVOF coatings. In

HVOF spraying, oxygen is present in the combustion gas

mixture and in the ambient air [36]. Therefore, carbon is

partially lost, and titanium oxide layers are formed on the

surface of the spray particles. During impingement of the

spray particles on the substrate, the oxides remain at the

boundaries of the spray splats. During wear, cracks can be

generated and grow at the brittle splat boundaries. These

cracks finally lead to the removal of the splats. Moreover,

the presence of oxides reduces the ductility of the overall

composite. Therefore, the coating exhibits a brittle wear

behaviour and almost all the material from the wear grooves

in the HVOF coatings is removed as debris (Fig. 8b). On the

contrary, for the VPS coating under the same wear

conditions, a considerable amount of the displaced material

from the wear grooves is ploughed aside the groove,

forming ridges, and thus, still remains in the coatings

(Fig. 8a). Hence, besides the delamination of material, the

overall lower ductility is another reason for the substantially

higher wear rate of the HVOF coatings compared to the

VPS coatings.

It should also be noted that the difference in the wear rate

between the nanocrystalline HVOF and VPS coating is

larger than that between the microcrystalline HVOF and

VPS coatings (Fig. 4). This difference seems to be related to

the oxide distribution in the HVOF coatings. The nano-

crystalline coating (Figs. 1c and 2c) shows some oxygen-

rich zones and thick bands of oxides, whereas in the

microcrystalline coatings oxides are thinner and more

homogeneously distributed (Figs. 1d and 2d). The smaller

oxygen-rich zones in the microcrystalline HVOF coating

contain fewer defects that can promote crack formation and

crack propagation compared to those in the nanocrystalline

coating. Therefore, less material delaminations occur,

resulting in a lower mass loss in the microcrystalline HVOF

coatings.

5. Summary and conclusions

Both nano- and microcrystalline HVOF coatings show

much lower wear resistance than the VPS coatings on the

basis of same feedstock powders, due to the presence of

oxide layers amongst the spray splats. The oxide layers

weaken the cohesion between splats and embrittle the

HVOF coating material, leading to a higher wear of these

coatings. The nanocrystalline HVOF coatings show a lower

wear resistance than the microcrystalline HVOF coatings,

due to more concentrated zones of oxides.

The microcrystalline VPS coatings have higher wear

resistance, if the stress exerted by individual abrasive

particles is low, such that the hard phase particles can carry

the load and support the counterbody. The wear mechanism

of the microcrystalline coatings undergoes a transition from

predominant removal of the binder phase to wear involving

hard phase fracture with increasing load of individual

abrasive particles. Under tougher wear conditions, the hard

phase particles are fragmented and removed together with

the binder matrix. In this case, the nanocrystalline VPS

coatings show superior wear resistance compared to the

microcrystalline coatings. In addition, the nanocrystalline

VPS coatings have also the advantage of showing a

smoother worn surface than the microcrystalline VPS

coatings. This is a significant advantage, if keeping a

smooth worn surface is more important than keeping the

clearance between wear parts low in the technical applica-

tion, e.g., rolls in paper industry.

The dominant abrasive wear mechanism of the VPS

sprayed nanocrystalline coatings can be described by cutting

and ploughing. The nanoscale hard phase particles and the

binder phase in the nanocrystalline coating are simulta-

neously displaced, forming ridges and wear debris, in which

the hard phase particles are still embedded in the binder

phase, as in the coating.


Acknowledgements

The support by W. Krommer and P. Heinrich from the

Linde AG, Unterschleißheim, Germany, in performing

HVOF experiments is greatly acknowledged. In addition,

the authors thank V. Borck and R. Henne, DLR, Stuttgart,

Germany, for their support in vacuum plasma spraying.

Thanks are due to M. Lembke and G. Favaro from CSM

Instruments, Switzerland, for the execution of scratch tests.

F. Felten from Technical University Hamburg, Germany, is

also acknowledged for his kind instruction during AFM

investigation.

This work was partly supported by the German Research

Society (DFG) within the collaborative research program

SFB 371 FMicromechanics of Multiphase Materials,_Project D3.

References

[1] L.-M. Berger, P. Vuoristo, T. Mantylat, W. Gruner, in: C. Coddet (Ed.),

Proc. of the 15th International Thermal Spray Conference, ASM

International, Materials Park, OH, USA, 1998, p. 75.

[2] B.Q. Wang, Wear 188 (1995) 40.

[3] M. Bjordal, E. Bardal, T. Rogne, Wear 186–187 (1995) 508.

[4] S. Berger, R. Porat, R. Rosen, Prog. Mater. Sci. 42 (1997) 311.

[5] J.M. Guilemany, J. Nutting, N.L. Isern, J. Therm. Spray Technol. 5

(1996) 483.

[6] S. Wirojanupatump, P.H. Shipway, D.G. McCartney, Wear 249 (2001)

829.

[7] R. Schwetzke, H. Kreye, J. Therm. Spray Technol. 8 (1999) 433.

[8] S. Zimmermann, H. Kreye, in: C.C. Berndt (Ed.), Thermal Spray: A

United Forum for Scientific and Technological Advances, ASM

International, Materials Park, OH, USA, 1996, p. 147.

[9] A. Bellosi, F. Monteverde, R. Calzavaroni, C. Zancolo, Int. J. Refract.

Met. Hard Mater. 19 (2001) 191.

[10] T. Rogne, J. Berget, T. Solem, in: German Welding Society (Ed.),

United Thermal Spray Conference, ASM Thermal Spray Society,

Materials Park, Dusseldorf, Germany, 1999, p. 487.

[11] T.H. Stenberg, K.J. Niemi, P.M.J. Vuoristo, J.E. Vuorinen, T.A.

Mantyla, in: C.C. Berndt (Ed.), Proc. ITSC ’95, ASM International,

Materials Park, OH, USA, 1995, p. 1145.

[12] L.M. Berger, M. Nebelung, P. Vuoristo, M. Heinonen, T. Mantyla, T.

Reinhardt, in: German Welding Society (Ed.), United Thermal Spray

Conference, ASM Thermal Spray Society, Materials Park, Dusseldorf,

Germany, 1999, p. 128.

[13] H. Doi, in: E.A. Almond, C.A. Brooks, R. Warren (Eds.), Proc. 2nd

Intern. Conf. Sci. Hard Materials, Inst. Phys. Conf. Ser., vol. 75, Adam

Hilger Ltd, Bristol and Boston, 1986, p. 489.

[14] I.M. Hutchings, Tribology: Friction and Wear of Engineering

Materials, Butterworth-Heinemann, Oxford, UK, 1992.

[15] K.-H. Zum Gahr, Microstructure and Wear of Materials, Elsevier

Science Publishers B.V., Netherlands, 1987.

[16] K. Jia, T.E. Fischer, B. Gallois, Nanostruct. Mater. 10 (1998) 875.

[17] K. Jia, T.E. Fischer, Wear 203–204 (1997) 310.

[18] W. Schlump, J. Willbrand, H. Grewe, Metallurgy 48 (1994) 34.

[19] B.H. Kear, P.R. Strutt, Nanostruct. Mater. 3 (1993) 19.

[20] B.H. Kear, R.K. Sadangi, M. Jain, R. Yao, Z. Kalman, G. Skandan,

W.E. Mayo, J. Therm. Spray Technol. 9 (2000) 399.

[21] Z. Fang, J.W. Eason, Int. J. Refract. Met. Hard Mater. 13 (1995)

297.

[22] K. Jia, T.E. Fischer, Wear 200 (1996) 206.

[23] S. Usmani, S. Sampath, D. Houck, D. Lee, Tribol. Trans. 40 (1997)

470.

[24] D.A. Stewart, P.H. Shipway, D.G. McCartney, Wear 225–229 (1999)

789.

[25] Y. Qiao, T.E. Fischer, A. Dent, Surf. Coat. Technol. 172 (2003) 24.

[26] Y. Qiao, Y. Liu, T.E. Fischer, J. Therm. Spray Technol. 10 (2001)

118.

[27] Y. Zhu, K. Yukimura, C. Ding, P. Zhang, Thin Solid Films 388 (2001)

277.

[28] N. Eigen, T. Klassen, E. Aust, R. Bormann, F. Gartner, Surf. Coat.

Technol. 159 (2005) 344.

[29] N. Eigen, T. Klassen, E. Aust, R. Bormann, F. Gartner, Mater. Sci.

Eng., A Struct. Mater.: Prop. Microstruct. Process. 356 (2003) 114.

[30] X. Qi, E. Aust, N. Eigen, F. Gartner, R. Bormann, Mat.-Wiss.

Werkstofftech. 35 (2004) 779.

[31] F. Gartner, H. Kreye, V. Borck, K. Krommer, in: C.C. Berndt

(Ed.), Proc. ITSC, ASM International, Materials Park, OH, USA,

2000, p. 463.

[32] F. Gartner, R. Bormann, T. Klassen, H. Kreye, N. Mitra, Mater. Sci.

Forum 343–346 (2000) 933.

[33] F. Gartner, H. Kreye, T. Klassen, R. Bormann, in: H. Dimigen (Ed.),

Proc. International Congress on Advanced Materials and Processes,

EUROMAT 99, Volume 11, Surface Engineering Held in Munchen,

Germany, September 27–30, 1999, WILEY-VCH, Weinheim, Ger-

many, 2000, p. 88.

[34] N. Eigen, Einsatzpotenzial des Hochenergiemahlens fur die Herstel-

lung nanokristalliner Werkstoffe, PhD thesis, Hamburg University of

Technology, Germany, 2004, p. 82.

[35] K. Schonert, in: K. Winnaker (Ed.), 4th edR Chemische Technologie,

vol. 1, Allgemeines, 1984, p. 80.

[36] H. Kreye, F. Gartner, A. Kirsten, R. Schwetzke, in: P. Heinrich (Ed.),

Proc. 5. Kolloqium Hochgeschwindigkeits-Flammspritzen, Erding,

Germany, November 16–17, 2000, Gemeinschaft Thermisches Sprit-

zen e.V, Unterschleißheim, Germany, 2000, p. 5.

Two-body abrasive wear of nano- and microcrystalline TiC–Ni-based thermal spray coatings

Documents

Transcript of Two-body abrasive wear of nano- and microcrystalline TiC–Ni-based thermal spray coatings