Spray Condition and Heat Treatment

8/12/2019 Spray Condition and Heat Treatment

1/6

Effect of Spray Condition and Heat Treatment on the Structure

and Adhesive Wear Properties of WC Cermet Coatings

Yasunari Ishikawa, Jin Kawakita and Seiji Kuroda

National Institute for Materials Science, 1-2-1 Sengen, Tsukuba 305-0045, Japan

An improved HVOF spray process called Gas-shrouded HVOF (GS-HVOF) has been developed over the past several years. By using an

extension nozzle at the exit of a commercial HVOF spray gun, GS-HVOF is capable of controlling the oxidation of sprayed materials during

flight as well as achieving higher velocity of sprayed particles. These features result in extremely dense and clean microstructure of the sprayed

coatings. The process has been successfully applied to corrosion resistant alloys such as SUS316L, HastelloyC, and alloy 625 as well as cermets

such as WC-Cr3C2-Ni. Wear properties of WC cermet coatings were measured by using a pin-on-disk wear tester. The specific wear rates of the

coatings prepared by the GS-HVOF with a reducing (fuel rich) flame were close to that of chrome plating. The wear amount of the heat-treated

GS-HVOF coatings could not be detected after testing. It is believed that transition to mild wear appeared early because of the increased surface

oxidation due to the heat treatment.

(Received March 25, 2005; Accepted May 9, 2005; Published July 15, 2005)

Keywords: high velocity oxygen fuel, gas-shroud, tungsten carbide cermet, wear property, heat treatment

1. Introduction

Hard chrome plating is widely used to deposit a film for

wear resistance. The effect of hexavalent chromium to the

environment and the operators health is a big problem.

Therefore, research for alternatives of chrome plating is

widely carried out in the world.1,2) HVOF-sprayed cermet

coating is one of such alternative candidates. HVOF process

owes both its heat source and acceleration force to a

supersonic jet flame made from a high-pressured mixture of

oxygen and fuel. This technique enables us to obtain sprayed

particles with a speed over 500 m/s and with a temperatureup to around 2000C. Such particles impinge onto a target

substrate often in the semi-molten state and pile up to form

coatings.

However, for some cermet powders, it is difficult to make a

dense coating. When one intends to decrease the coating

porosity by usual HVOF equipment, such spray conditions

tend to increase the degree of oxidation of the sprayed

coating. The increase of oxidation can be reasons of lowering

abrasive wear resistance of the sprayed coating. Therefore,

we have developed a gas-shroud attached HVOF (GS-

HVOF), which can introduce an inert gas at the exit of the

barrel into the shroud attachment (Fig. 1). With thismodification, it is possible to suppress the oxidation of

sprayed particles while raising the velocity of sprayed

particles.3,4)

In our previous work,5) by using the gas-shroud attach-

ment, corrosion resistance of sprayed coatings increased with

increase of density. Wear resistance of the coatings against

abrasion increased with increase of the hardness of the

coating. These classes of coatings are widely used as wear

resistant coating for the steel making rolls and the paper-

making rolls in Japan. The coatings used in such applications

are rubbed repeatedly against steel or paper. In this case, the

wear mode is dominated by the adhesive wear.

When metals slide against each other, it is generally

accepted that severe-mild wear transition occurs. In other

words, adhesive wear changes from the high wear rate

condition (severe wear) to the low wear rate condition (mild

wear).6)

In the early stage of adhesive wear, an oxide layer on the

surface is destroyed and the exposed fresh surfaces adherewith each other and are worn apart repeatedly. The surface is

worn severely and the large metallic wear debris is generated

(0.10.2 mm). The specific wear rate lies in the range from

105 to 106 mm2/N. This is called the severe wear. Once

the worn surface is covered with the oxidized wear debris,

specific wear rate decreases by a factor of 1/100 and small

dark wear debris is generated. The specific wear rate becomes

below 108 mm2/N. This is called the mild wear. The wear

rate transition like this phenomenon is called the severe-mild

wear transition.7)

A number of papers have been published about the wear

properties of thermal spray coatings, mostly about theabrasive wear and erosion wear.8,9) In some researches of

adhesive wear, the wear property of thermal spray coatings is

only evaluated for the relationship between the specific wear

rate (volume wear loss) and the mechanical properties such as

hardness, which depend on the porosity and the micro-

structure of the coatings.1012)

In this study, we investigate the adhesive wear property of

HVOF sprayed cermet coatings by using a pin-on-disk wear

tester. The effect of spray condition and heat treatment on the

structure and wear properties such as the specific wear rate

and the severe-mild wear transition are discussed in detail.

2. Method

Coatings were prepared with a commercial HVOF

(JP5000, Praxair TAFA) and the GS-HVOF thermal sprayFig. 1 Schematic representation of HVOF with the gas shroud attachment.

Materials Transactions, Vol. 46, No. 7 (2005) pp. 1671 to 1676#2005 The Japan Institute of Metals


2/6

equipment, which use kerosene and oxygen to generate a

supersonic combustion flame. A commercially available

cermet powder (SURPREX W2007L, WC/20Cr3C2/7Ni,

FUJIMI Inc.) with size distribution from 15 to 53 mm was

used as the feedstock. The substrate was JIS SS400 low

carbon steel and its surface was blasted by alumina grit and

degreased by ultrasonic cleaning in acetone. Thickness of all

the coatings was aimed at about 300 mm. The basic operating

conditions are listed in Table 1. Table 2 shows detailed

spraying conditions used for spraying, i.e., fuel flow rate,

oxygen flow rate, the fuel to oxygen ratio and combustion

pressure Pc. F/O is the fuel to oxygen ratio, where F/O 1corresponds to the stoichiometoric mixture ratio for complete

combustion. Coatings were prepared by the oxidizing flame,

the reducing flame and with the gas shroud, which operates

with the reducing flame.

We measured the in-flight velocity and temperature of

sprayed particles by an instrument called the in-flight thermal

sprayed particle analyzer (DPV-2000, TECNAR Co.) at the

substrate position. Its principle and mechanism were describ-

ed in detail elsewhere.13) This equipment utilizes the thermal

radiation from the in-flight sprayed particles. The image of a

hot particle is formed on a pair of vertical slits by means of a

lens. Temperature is determined by the two-color pyrometry.Velocity is determined by dividing the interval (160 mm) of

the two slits by the time between two radiation signals

detected when the image of one particle passes in front of the

slits.

Wear property of the coatings was evaluated by using a

pin-on-disk type wear tester (Fig. 2). Coated disk surfaces

were polished with diamond grinding, and finished with

buffing so that their surface roughness had a Ra 0.01

0.02 mm. Furthermore, some GS-HVOF coatings were heat

treated in air atmosphere before wear testing. The samples

were heated at a rate of 20C/min up to 500C and held at the

temperature for 30 h.

The diameter of the iron pin (Fe, >99:5%) used in the

experiment was 4 mm. After applying ultrasonic cleaning to

the pin and the disk in acetone for 1 h, they were fixed to the

test apparatus. The pitch circle diameter (P.C.D.) was 20 mm.

In this experiment, the sliding velocity was fixed at 0.25 m/s

for the entire test. The load between the pin and the disk was

10 N in the tests. The wear amount of coatings was measured

after 3000 m sliding. The specific wear rates are the averages

of three measurements on each sample. The displacement ofthe top surface of the pin was measured by using a gap sensor

during the wear test. Wear rate was defined as the volume

loss per unit load and per unit sliding distance. For

comparison, the wear amount of chrome plating was also

measured. Coatings, sliding surfaces and wear debris were

observed by optical microscope and were analyzed by SEM

(EDX), XRD techniques.

3. Results and Discussion

Figure 3 shows spray particles in-flight average velocity

and temperature measured by the in-flight thermal sprayed

particle analyzer. Spray particles in-flight velocity increasedfrom 680 to 820 m/s as it went from the oxidizing flame to

the gas shroud condition. By using the gas shroud equipment,

the particle temperature fell by 150 degree, whereas the

particles velocity increased. Each error bar shows the

maximum value and minimum value obtained by all data.

The hardness on the cross section of coatings was

measured by a Vickers hardness tester at a 3 N load. Figure 4

shows the relationship between the hardness of coatings and

the flight velocity of spray particles. The hardness values are

the averages of ten measurements on each sample. Coatings

hardened with the increase of spray particles in-flight

velocity. The coating hardness did not depend on the spray

Table 1 Basic operating conditions.

Barrel length mm 150

Powder feed rate g/min 70

Torch velocity mm/s 700

Spray distance mm 300

Powder feed gas Nitrogen

Coating thickness mm 300

Table 2 Spraying conditions.

Fuel Oxygen F/O Pc

Unit dm3/min dm3/min MPa

Oxidizing flame 0.32 897 0.73 0.63

Reducing flame 0.44 826 1.09 0.67

Gas shroud 0.44 826 1.09 0.67

Load

Pin

Substrate

Gap

sensor

Coating

d

Disk:30 30mm(Ra0.01)

Pin :4(Fe, Hv 138)

Load: 10N

Velocity:0.25m/s

P.C.D.: 20mm

Distance:3000m

Fig. 2 Pin-on-disk type wear tester.

600

650

700

750

800

850

Oxidizing flame Reducing flame Gas shroud

In-flightsprayparticlesvelocity,V/m

s-1

1750

1800

1850

1900

1950

2000

In-flightsprayparticlestemperature,T/C

Velocity

Temperature

Fig. 3 In-flight particles velocity and temperature. Each error bar shows

the maximum value and minimum value obtained by all data.

1672 Y. Ishikawa, J. Kawakita and S. Kuroda


3/6

particles in-flight temperature. Lidonget al.investigated the

influence of the spray particles in-flight velocity and the

spray particles in-flight temperature exerted on hardness of

WC-CoCr coating on mild steel St 37 made by HVOF. The

coatings hardness increased with increasing the spray

particles in-flight velocity and the spray particles in-flight

temperature.14) From these results, we consider that the

coating hardness was increased with increasing of the work

hardening or the residual stress that were increased by the

increase of the spray particles in-flight velocity.

The coating made by the GS-HVOF process was heat-

treated at 500C for 30 h. Its hardness was not changed by the

heat-treatment. This result appears to be contrary to the reportby Okada et al.15) They investigated the effect of heat

treatment on the hardness of WC-27NiCr coating on SUS403

made by HVOF. Coatings were heated to either 300, 400,

500C and held at those temperature between 1 h and 45 h.

The hardness of all coatings was increased by heat treatment.

The hardness of the coating heat treated at 500C showed the

greatest increase,i.e.the hardness was increased from Hv750

to HV1100. The amount of voids in the coatings wascalculated by the image processing of SEM image. The

increase of the hardness was attributed to the decrease of

voids in the coatings. Since the coating by GS-HVOF in the

present study has essentially no voids and pores in the as

spray condition. Therefore, the coating hardness was not

increased by the heat-treatment.

We observed the cross section of the coatings by using

SEM (Fig. 5). Boundaries among the particles were observed

in the coating prepared with the oxidizing flame. Pores and

voids are also visible on the cross section. Lamellar structure

was slightly revealed in the cross section of the coating

prepared by the reducing flame. The spray particles boundary(lamellar structure) decrease gradually as it goes from the

oxidizing flame to the reducing flame. By using the gas

shroud equipment, boundaries are hardly observed and the

lamellar structure cannot be identified. This result indicates

the improvement in density of coatings by the gas shroud

attachment.

The crystal structures of the coating were characterized

from the surface by XRD (Fig. 6). We confirmed the

existence of Cr3C2 and Cr2O3 peak. The ratio of Cr2O3

Cr3C2 was calculated by using the intensity of the main peak

of each material (Fig. 7). The heat-treated coating showed

the highest oxidation degree in all the specimens. During heat

treatment (at high enough temperature) oxygen combineswith Cr in the coating to form Cr2O3.

16) The XRD diffraction

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

2000

600 650 700 750 800 850 900

In-flight velocity of particles, V/m?s-1

Vickershardness,Hv0.3

Fig. 4 The relationship between the coating hardness and In-flight

particles velocity. The hardness values are the averages of ten measure-

ments on each sample. Each error bar shows the maximum value and

minimum value obtained by all data.

5mVoid

WC

CrC

Ni

Boundary

5m

ReducingFlame

OxidizingFlame

5m

5m

Gas

Shroud

HeatTreatment

CrC

WC

CrC

Boundary

WC

Ni

Ni

CrC

WC

Ni

Fig. 5 Cross section SEM image of coatings. White: WC, Light gray: Ni, Black: Cr3C2.

Effect of Spray Condition and Heat Treatment on the Structure and Adhesive Wear Properties of WC Cermet Coatings 1673


4/6

spectrum of the heat-treated coating shows peaks indexed to

NiWO4. However, after polishing 50 mm by an abrasive

paper, the oxidation degree was decreased and XRD peaks of

inside of coating was similar to that of gas-shroud coatings.

The coating by the reducing flame showed the highest

oxidation degree among the untreated specimens.This result was in agreement with Kreyes research, which

showed the effect of F/O ratio on the oxide content of the

coatings.17) The formation of oxide during the thermal spray

process can be categorized into three regions.18) Region 1 is

in the jet core, region 2 is in the atmosphere from the outside

of jet flame to the substrate and region 3 is on the substrate.

The spray particles in-flight temperature with the reducing

flame was little higher than that by the other flames. The

oxygen content of the coatings depends not only by the heat

of jet flame, but also by the substrate temperature. Since the

reducing flame burns further at the outside of the nozzle exit,

the substrate temperature should be higher than other

combustion flames. Therefore, the coating by the reducing

flame showed the highest oxidation degree among the

untreated specimens.

Figures 8 and 9 show the specific wear rates of the coatings

and the iron pin. Each error bar shows the maximum value

and minimum value of the specific wear rate obtained by 3runs. The specific wear rate of the coating prepared by the

oxidizing flame was greater than the other coatings. The

specific wear rate of the coating prepared by the reducing

flame and GS-HVOF flame was in the order of109 mm2/N

and was close to that of chrome plating. The specific wear

rate of the heat-treated coating is shown as zero, because the

wear amount of the heat-treated coating could not be detected

after the wear testing. Even though hardness is not different

between the heat-treated GS coating and the untreated GS

coating, the specific wear rate of the heat-treated GS coating

is much lower than that of the untreated GS coating. From the

result of Fig. 7, we consider that the adhesive wear propertywas influenced by the existence of oxides.

The specific wear rate of the iron pin was in the order of

108 mm2/N except for the combination of the iron pin and

the coating prepared with the oxidizing flame. Although the

wear amount of the heat-treated coating could not be detected

after the wear testing, the specific wear rate of the pin showed

a value in the order of 108 mm2/N, which is little higher

than that of the iron pin against the coatings prepared by the

GS-HVOF.

With the comparison of these wear behaviour, it is obvious

that the specific wear rate of the combination of the oxidizing

flame coating was the highest among the tested combinations.

However, even in this case the specific wear rate of the pin

was just above 107 mm2/N. From such low wear rates of

both the pin and the coatings, it is likely that the wear of all

the combinations was in the mild wear mode after 3000 m

WC, W2C, WO3, Cr3C2, Cr2O3, NiO, NiWO4

30 40 50 60 70 80 90

30 40 50 60 70 80 90

-Heat treatment

-Heat treatment

(50mm inside)

30 40 50 60 70 80 90

-Oxidizing flame

-Reducing flame

-Gas shroud

2

Fig. 6 XRD patterns of coatings.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Oxydizing flame Reducing flame GS Heat treatment

TherelativeintensityofCr2O3/Cr3C2

Fig. 7 Oxidation degree during spraying in terms of the XRD intensity

ratio of Cr2O3/Cr3C2.

0

0.5

1

1.5

2

2.5

Oxidizing

flame

Reducing

flame

Gas shroud Chrome

plating

Heat treated

Specificwearrate,W/10-8mm

2

N-1

Fig. 8 Specific wear rate of coatings.

0

5

10

15

20

25

Oxidizing

flame

Reducing flame Gas shroud Chrome plating Heat treated

Specificwearrate,W/10-8mm

2

N-1

Fig. 9 Specific wear rate of iron pin.



5/6

sliding.

To examine the severe to mild wear transition, we recorded

the displacement of the pin head as shown in Fig. 2.

Figure 10 shows the obtained typical displacement curves.

The inclination of specific wear rate slightly changes by eachmeasurement. Since the wear amount of coatings is very low,

the displacement curve of d may be regarded as the wear

behaviour of iron pin.

In our experiments, only the coating made by the oxidizing

flame showed the severe-mild wear transition at 50 m,

whereas the other coatings did not show the severe-mild

wear transition. Since the coating made by the oxidizing

flame has the highest porosity and the particles boundary is

more obvious larger wear debris might be generated easily at

early wear stage, because of detachment of loosely bonded

sprayed particles. Since small oxide wear debris is generated

by crushing such larger wear debris, the wear behaviour of a

coatings changes from severe wear to mild wear. Therefore,we consider that only the coating made by the oxidizing

flame showed the severe-mild wear transition.

Furthermore, we observed the wear debris of the latter

coatings without severe-mild wear transition during wear

testing. Dark-colored and small (15mm) wear debris was

generated on the surface at an early wear stage. Wear tracks

and the iron oxide existed on the surface of the coating after

wear testing. As latter coatings showed the low specific wear

rate (108 mm2/N), these results indicated that the latter

coatings realized the mild wear condition at a very early stage

of testing.

In previous work,19)

we found that the abrasive wearresistance of the coatings depended on the hardness of the

coatings. However, in the adhesive wear mode such as the

pin-on-disk type wear test in this study, the wear rate does not

simply depend on the hardness of coatings but also on the

chemical composition of the coating surface. XRD analysis

of the heat-treated coating from surface indicated the peaks

of NiWO4, WO3. It seems that the adhesive wear property

was influenced by the existence of these oxides.

Kato investigated the influences of oxide particles supplied

to the rubbing surface during sliding on the severemild

wear transition by using a pin-on-disc wear tester.20) As a

result, oxide particles were found to play an important role in

reducing the amount of wear. The sever-mild wear transition

distance was decreased by decreasing the size of the supplied

oxide particles.

He surmised that this effect is caused by the formation of

the oxide layers on the rubbing surfaces.

In addition, Picaset al.reported that the specific wear rate

of thermal sprayed nanocrystalline CrC-NiCr coatings was

decreased by the decrease of the spray powder size by using a

pin-on-disk wear tester.21)

Since the original particle size of our WC cermet powder

feedstock is very small (about 12 mm) and these smallparticles are partially oxidized from surface by the spray

flame, it is expected that fine oxide particles were dispersed

in the coatings by thermal spray technique. The thermal spray

coatings might easily generate the small oxide wear debris by

wear than coatings made by other processes. It is well known

that the amount of oxide in the thermal spray coatings is

increased by the decrease of the spray powder size.22)

As the oxidation of such thermal spray coating is promoted

by the heat treatment in air especially on the coating surface,

the small oxide wear debris could be obtained from an early

wear stage. We consider that this behaviour is similar to the

effect of shortening the severe-mild wear transition distanceby supplying the oxide particles to the rubbing surface during

sliding. This is one explanation why the severe-mild wear

transition was accelerated by heat treatment but more

detailed study is needed.

4. Conclusion

WC/20Cr3C2/7Ni cermet powder was HVOF sprayed

with combustion flames with different fuel-oxygen mixture

ratio. GS-HVOF was also employed to obtain dense coatings

with minimum oxidation. The results show that the coating

obtained by the fuel-rich flame was oxidized most. Heat

treatment in air increased the oxidation on the coating surfacesignificantly.

Adhesive wear resistance of the coatings was evaluated by

the pin-on-disk wear tester. The result has demonstrated the

following.

(1) The specific wear rates of the cermet coatings were in

the order of 109 mm2/N, an order of magnitude

smaller than that of the iron pin. Only the coating

sprayed with the oxidizing flame exhibited the severe-

mild wear transition and the specific wear rate of the pin

was slightly above107 mm2/N.

(2) Although the hardness of the coatings was not different

between the heat-treated GS coating and the untreatedGS coating, the specific wear rate of the heat-treated GS

coating was much lower to the extent that it could not be

measured. This was attributed to the effect of oxides

formed by the heat treatment, which should promote the

severe-mild wear transition by forming fine oxide

particles in between the sliding surfaces.

This might have a practical significance it is important

for high wear resistance materials that mild wear

appears early.

Acknowledgement

We would like to thank Dr. T. Itsukaichi and Mr. S. Osawa

of Fujimi Inc. for supplying the cermet powder. Careful

preparation of the coating samples by Mr. M. Komatsu and

Mr. N. Kakeya of NIMS are also appreciated.

0

50

100

0 100 200 300 400 500

0

50

100

150

200

250

300

0 500 1000 1500 2000 2500 3000

Sliding distance, l/m

Displacementofironpin,d/m

Gas shroud

Reducing flame

Oxidizing flame

Heat treatment

Chrome plating

Fig. 10 The displacement of iron pin.

Effect of Spray Condition and Heat Treatment on the Structure and Adhesive Wear Properties of WC Cermet Coatings 1675


6/6

REFERENCES

1) B. Sartwell, K. Legg and J. Sauer: Joint Test Report, (Validation of

WC/Co HVOF Thermal Spray Coatings as a Replacement for Hard

Chrome Plating On Aircraft Landing Gear, PART I: MATERIALS

TESTING, U.S. DEPARTMENT OF DEFENSE, 2002).

2) J. G. Legoux:Proc. 7th Workshop on the Ultra-Steel, (NIMS, 2003)

pp. 124133.3) J. Kawakita, S. Kuroda, T. Fukushima and T. Kodama: Sci. Technol.

Advance Mater.4 (2003) 281289.

4) J. Kawakita and S. Kuroda: Rust Prevention & Corrosion Control48

(2004) 287294.

5) Y. Ishikawa, J. Kawakita and S. Kuroda: Proc. 1st Tsukuba Interna-

tional Coating Symposium, (NIMS, 2004) pp. 4041.

6) W. Hirst and J. K. Lancaster: Journal of Applied Physics 27 (1956)

10571065.

7) T. Sasada, Y. Tsukamoto and K. Mabuchi:Biotribology, (Tokyo, 1990)

pp. 20.

8) D. A. Stewart, P. H. Shipway and D. G. McCartney: Wear 225229

(1999) 789798.

9) D. Toma, W. Brandl and G. Marginean: Surf. Coatings Technol. 138

(2001) 149158.

10) L. Jacobs, M. M. Hyland and M. De Bonte: J. Thermal Spray Technol.

8 (1999) 125132.

11) L. Zhao, M. Maurer, F. Fischer, R. Dicks and E. Lugscheider: Wear257

(2004) 4146.

12) J. Wang, K. Li, D. Shu, X. He, B. Sun, Q. Guo, M. Nishio and

H. Ogawa: Mater. Sci. Eng. A 371 (2004) 187192.

13) H. Yamada, S. Kuroda, T. Fukushima and H. Yumoto: Proc. ITSC2001,

ed. by C. C. Berndt, (ASM International, Singapore, 2001) pp. 797

803.14) L. Zhao, M. Maurer, F. Fischer, R. Dicks and E. Lugscheider: Wear257

(2003) 4146.

15) R. Okada and M. Yamada: J. Japan Inst. Metals58 (1994) 763767.

16) J. M. Guilemany, J. M. Miguel, S. Vizcaino, C. Lorenzana, J. Delgado

and J. Sanchez: Surf. Coatings Technol.157 (2002) 207213.

17) K. Dobler, H. Kreye and R. Schwetzke: J. Thermal Spray Technol.9

(2000) 407413.

18) C. M. Hackett and G. S. Settles:Proc. 8th NTSC, ed. by C. Berndt and

S. Sampath (ASM International, Texas, 1995) pp. 2129.

19) Y. Ishikawa, J. Kawakita and S. Kuroda: J. Thermal Spray Technol., in

print.

20) H. Kato: Wear255 (2003) 426429.

21) J. A. Picas, A. Forna, A. Igartua and G. Mendoza: Surf. Coatings

Technol. 174175 (2003) 10951100.

22) C.-J. Li and W.-Y. Li: Surf. Coatings Technol.162 (2003) 3141.


Documents

Spray Condition and Heat Treatment