FRACTOGRAPHICAL ANALYSIS OF WORN …metal2013.tanger.cz/files/proceedings/metal_08/Lists/...METAL 2008 13. –15. 5. 2008, Hradec nad Moravicí _____ 1 FRACTOGRAPHICAL ANALYSIS OF

METAL 2008 13. –15. 5. 2008, Hradec nad Moravicí

___________________________________________________________________________

1

FRACTOGRAPHICAL ANALYSIS OF WORN SURFACES OF

DUPLEX-COATED P/M VANADIS 6 LEDEBURITIC STEEL

Peter Jurčia

Mária Hudákováb

Jaroslav Maixnerc

aČVUT, Fakulta strojní, Karlovo nám. 13, 121 35 Praha 2, p.jurci @seznam.cz

bMtF STU Trnava, Pavlínská 16, 917 24 Trnava, SR cVŠCHT Praha, Technická 6, 160 00 Praha 6

ABSTRACT

Duplex-coated P/M made Vanadis 6 ledeburitic steel was subjected to the wear testing.

After the testing, worn specimens were analysed by the fractography and connected methods

in order to determine what happened during the sliding. It was found that the samples differ

one from each other in the adhesion of PVD-overlay on nitrided substrate. The specimens

with good adhesion have shown a critical load of more than 130 N, wear of specimens was

slow, wear of the counterparts very rapid and the worn surface manifested the occurrence of

slightly worn CrN-overlay. On the other hand, if the coating was damaged, the wear rate of

the specimen increased rapidly and the worn surface showed the symptoms of massive

cracking and total collapse of the coating. In addition, the samples exhibited differences in

chemistry and phase constitution of PVD-layers.

1. INTRODUCTION

Hard coatings made by various physical vapour deposition (PVD) methods are widely

used to modify the surface properties of tool steels. They can be deposited in large scale of

chemical composition and mechanical properties, at low temperatures and without an

influence on core properties of tools. Common disadvantage of the PVD - coatings is the

adhesion on steel substrate since they differ from steels considerable in terms of mechanical

and physical properties. The formation of diffusion bonding is impossible because of low

process temperature. Therefore, the main experimental effort tends to the development of

layers with properties as similar to the substrate as possible. The second main trend is to

modify the substrate to increase the load carrying capacity and prevent plastic deformation.

The heat treatment affects the Young modulus of the steels in a very limited range. On

the other hand, the hardness can be increased very significantly. For the steel Vanadis 6, the

hardness in as-delivered state is of about 255 HB (23 HRC) and after heat treatment the

hardness of 62 HRC (750 HV) can easily be achieved. An additional hardness increase (and

also the hardness/Young modulus ratio) can be achieved with nitriding. For the Vanadis 6

steel a hardness of 1100 HV can be achieved by plasma nitriding [1,2].

Thin CrN coatings have been developed over the past 25 years [3]. They have early

gained an interest and popularity in a variety of industrial applications due to wear and

corrosion resistance [4], also at a high temperature [5], and good cutting properties in copper

machining [6], aluminium die casting and forming [5] and wood processing [7].

CrN – layers can be produced in a very wide range of chemistry, phase constitution and

properties. These characteristics influence, among others, also the adhesion on the substrate.

Generally, a good adhesion can be achieved when coating the ledeburitic steels or cemented

carbides [8-11]. Adhesion also increases as the portion of the phase CrN decreases and Cr2N

and/or (Cr) solid solution increases [12-14]. Also the annealing of the specimens after

deposition acts in favour of better adhesion [8,14].


___________________________________________________________________________

2

The microhardness of coating is ranges normally between 1500 and 3000 HV [6,15,16].

Mayrhofer [15] suggested that the hardness is a function of the grain size of CrN and that

there is the optimal grain size (nm) at which the hardness is maximal. Annealing of coated

materials in order to optimize the stress situation leads to decrease of hardness [8].

Also the Young modulus of CrN coatings can be modified in a wide range, commonly

from 188 to 400 GPa, depending on the substrate temperature and negative substrate bias,

respectively [13,16]. Interesting results presented Mercs et al. [16] for reactively sputter

deposited films on the M2 type ledeburitic steel. The Young modulus was 240 – 320 GPa,

directly proportioned to the hardness of coating. The most important information is, however,

that some of investigations indicated that the CrN-film can be deposited with the Young

modulus not much different from the steel substrate.

Formation of CrN – layers is connected with high compressive residual stresses in these

films. Lamastra [9] reported residual stresses ranging between -3.8 and -4.1 GPa for the film

deposited on the H13 hot work steel and -4.9 - 5.1 GPa for that formed on K340 ledeburitic

steel. For the M2 – ledeburitic steel, Djouadi [17] reported the compressive stresses only 1.5 –

2.5 GPa. From this reason, Odén [8] and Broszeit [14] independently suggested the heat

treatment (annealing) to modify the stress situation in CrN-layers. Broszeit [14] observed that

the compressive stresses in CrN-film decreased from 5797 MPa after annealing at 200oC to

1953 MPa after annealing at 500 oC. Odén [8] reported similar stress behaviour for the CrN –

layer deposited on PM tool steel - the stress was reduced to 1.6 GPa after annealing at 300 oC.

Most part of coatings shows a texture with a dominant orientation of CrN (111) and

(200), respectively. For the films deposited on ledeburitic steel substrates, also a preferred

orientation of (220) was reported [8,10-12]. In addition, some investigation indicated that the

orientation (texture) can closely be related to actual processing parameters [18].

2. EXPERIMENTAL

Specimens 70 x 18 x 8 mm made from the steel Vanadis 6 (2.1 %C, 7 %Cr, 6 %V, Fe

bal.) were used as a substrate. They were mechanically ground and vacuum heat treated to a

hardness of 60 HRC. Then, they were polished to a roughness Ra = 0.06 – 0.07 µm. Plasma

nitriding was performed in the device RUBIG Micropuls – Plasmatechnik®

at temperatures of

500 and 530 oC and processing times 60 - 120 min. After that, the surface roughness increased

to Ra = 0.09 – 0.1 µm. CrN – layers of thickness 2 or 5 µm were deposited by chromium

evaporation in a Balzers BAI 730M thermionic arc sputtering device at a constant nitrogen

partial pressure and temperature of 450 oC. Adhesion of layers was examined with the scratch

test. The critical load Lc was determined as the first adhesion failure, by identification of

acoustic emission echo as well as by viewing of micrographs of scratches. Microstructural

examinations were performed with scanning electron microscopy. Glow discharge optical

emission spectroscopy (GDEOS) was used to determine the elements depth profiles. Phase

constitution of CrN-layers was identified with X-ray diffraction. Wear testing was carried out

without lubrication using a ring-on-plate configuration of the counterpart and tested specimen,

at loads of 50 N and 150 N. Hollow cylinders from the ball bearing steel 100Cr6 heat treated

to 60 HRC were used as counterparts. The total distance (sliding in a combination with slow

rotation) was 10 km and the weight loss was determined also after 1, 2.5 a 5 km of sliding.

3. RESULTS, DISCUSSION

In the previous paper [19], the best nitriding parameters with respect to the properties

were found to be: temperature of 530 oC, process time of 120 min and reactive atmosphere

N2:H2 = 1:3. Tribological tests showed a general tendency for improvement of the wear

resistance but at a large scattering of measured values. Therefore, it was decided to examine

also other combinations of nitriding parameters, to reduce the range of dispersion of results.


___________________________________________________________________________

3

Figs. 1, 2 – SEM micrographs of duplex layers. Left – plasma nitriding 500 oC/60 min. + CrN

2 µm, right - plasma nitriding 530 oC/60 min. + CrN 5 µm

Microstructure of duplex-layered Vanadis 6 steel is in Fig. 1 and 2. On the surface,

CrN-layers of thicknesses of 2 and 5 µm are visible. Nitrided inter-layer made at 500 oC for

60 min. has a thickness of 25 µm, that formed at 530 oC for 120 min. of about 40 µm.

Nitrided regions differ from the core material mainly in the topography caused by etching.

Also a quasicontinuous nitride network was found at the grain boundaries, especially in the

case of region formed at 530 oC for 120 min, Fig. 2. No inhomogenities were found in the

surface regions and the PVD-layer was free of cracks and pores.

All the newly used combinations of plasma nitriding and CrN-layering produced a

minimal adhesion of approx. 50 N, Tab. 1. This indicates that the pre-treatment ahead the

plasma nitriding is a promising method for increasing the adhesion of PVD CrN-coating. On

the other hand, however, the adhesion values exhibited again a large dispersion even for the

specimen processed in the same conditions.

Results of measurements of the wear resistance for the load of 50 N, Fig. 3, demonstrate

good wear behaviour of the duplex-treated material in comparison to the material without

surface processing, only nitrided or only CrN-layered [4]. When compared to specimens pre-

nitrided at 530 oC for 120 min and then CrN-layered [20], the results of the tests were not

sufficiently reliable. However, it is clear that the pre-nitriding at 500 oC for 120 min. is rather

more efficient for the wear behaviour of the duplex layer than the pre-nitriding 530 oC/60

min. Also, to the opposite of previous tests [1,20] it was confirmed that thinner CrN-layer had

a better wear resistance. This wear behaviour can be expected since internal stresses, often

leading to the coating damage increase with the coating thickness [21].

Processing Lc [N]

CrN 2 µm 18

CrN 5 µm 26

Nitriding 530 oC/120 min + CrN 2 µm 44

Nitriding 530 oC/120 min + CrN 5 µm 90

Nitriding 500 oC/120 min + CrN 2 µm 50, not identified

Nitriding 500 oC/120 min + CrN 5 µm 55, 148

Nitriding 530 oC/60 min + CrN 2 µm 46,135

Nitriding 530 oC/60 min + CrN 5 µm 47, 52

Tab. 1 – Measurements of adhesion with scratch test. Yellow lines – results in [4].

1 17 µm 2 17 µm


___________________________________________________________________________

4

Measurements gave lower weight loss for the counterparts after sliding with the CrN-

layer with a thickness of 2 µm (Fig. 4). No strong changes in weight loss were found which

would indicate that no delamination of coatings occurred during the sliding test. It is

concluded that the cause of different wear was the differences in the friction coefficient (Fig.

5) – higher friction coefficient induced a rather higher wear, (Fig. 3).

Fig. 3 – Weight loss of specimens as a function of sliding distance and surface treatment, load

of 50 N.

Fig. 4 – Weight loss of counterparts as a function of sliding distance and surface treatment,

load of 50 N.

The use of a load of 150 N led to highlighted differences in wear behaviour between the

specimens processed by various surfacing methods (Fig. 6) but also between specimens

processed in the same batch (Fig. 8). However, the general tendency of improved wear

resistance after plasma nitriding and also the fact, that there is rather no general tendency in

wear behaviour as a function of nitriding parameters, was retained. The assessment of other


___________________________________________________________________________

5

effects on wear resistance became more difficult due to sudden changes on the sliding surface

due to the PVD-layer delamination. Therefore, the mean values from Fig. 6 do not represent

the wear behaviour accurately because of large dispersion of experimental data.

Fig. 5 – Friction coefficient as a function of sliding distance, load of 50 N

Fig. 6 – Weight loss of specimens as a function of sliding distance and surface treatment, load

of 150 N

Also the figures of weight loss of counterparts exhibited a large scattering. The mean

values in Fig. 7 give only partial information on the process during the wear testing. To

understand these differences in wear behaviour, selected specimens with good (71) and worse

(72) adhesion were subjected to more detailed checking.

In the cases of absence of delamination, the weight loss of the specimen was minimal,

the counterpart undergoes an intensive wear (Fig. 8) and the friction coefficient increased

slowly. On the other hand, if the adhesion of the CrN-layer was poor, the wear rate on the

specimen´s side changed suddenly from low to a very high, and the weight loss of the

counterpart became minimal. Simultaneously, also the friction coefficient decreased (Fig. 9)

as a result of decreased contact between the specimen and counterpart.

These observations correspond well with the scratch-test measurements although the

type of contact is different. As shown already, the adhesion of 50 N, which is the minimal

acceptable value for various industrial applications, was achieved for all the specimens. In


___________________________________________________________________________

6

some cases, however, the critical load for first flaking exceeded substantially 100 N.

Compared to many published data from the investigations of CrN coatings made on the

surfaces of either ledeburitic steels [8,12] or cemented carbides [6,7] it is apparently shown

that the pre-treatment via plasma nitriding can have a beneficial effect on adhesion. Large

differences in the wear resistance found at a load of 150 N can be attributed to the scattering

of adhesion values (ranged between 50 and 148 N) determined at the scratch test.

Fig. 7 – Weight loss of counterparts as a function of sliding distance and surface processing of

the specimen, load 150 N.

Fig. 8 – Wear of specimen and counterpart. Good (yellow) and poor (red) adhesion of layer.

Figure 10 shows the surface of specimen with an excellent PVD layer adhesion. The

surface shows two areas of different micromorphology. The first area (part A) is typical for

the practically undamaged layer and only the friction producing parallel tracks in the sliding

direction occurred. On the second, “B” area, a slightly deformed and cracked coating without

clear marks of delamination is found. The preservation of the coating on the surface inspite of

the high contact pressure confirms its excellent adhesion.


___________________________________________________________________________

7

Fig. 9 – Friction coefficient for the specimen with good adhesion (yellow) and poor adhesion

(red) of the CrN coating

Figs. 10,11 – Worn surface of the specimen with a good adhesion of CrN-layer (yellow) and

poor adhesion (red)

In the case of poor adhesion of the coating (Fig. 11), the wear process was probably

equal as in the case of a good adhesion (Fig. 10) up to occurrence of cracking of the layer at

various places, mostly however at the boundary between the worn and the unaffected parts of

the specimen (A). From this moment on, the friction coefficient started to decrease (Fig. 9).

The degradation of the layer continued with crumbling dominantly in the vicinity of primary

cracks (B), up to the total material removal in selected areas (C). Removed fragments of

coating accelerated the weight loss on the specimen´s side due to the removal itself, while, on

the counterpart´s side, the weight loss was practically stopped due to minimizing of the

sliding contact and lowering of the friction coefficient (Fig. 9).

Specimens with good (71) and worse (72) wear behaviour differ also in the chemistry

and phase constitution of PVD-layers. The layer on the specimen 71 contains less Cr and

more nitrogen than that on the specimen 72 (Figs. 12,13). The difference is of 1.5 – 2%,

depending on the depth. At the same time, the layer on the specimen 71 consists mainly of the

Cr2N – compound, while, in the layer on the specimen 72, the content of chromium solid

solution is significantly increased (Fig. 14). This observation is in a good agreement with the

data reported by several authors [13,14] where the higher chromium content containing layers

have generally shown better adhesion on the steel substrates. It can also be assumed that the

A

B S

lidin

g d

irection

A

B

C

Slid

ing

directio

n


___________________________________________________________________________

8

hardness of CrN film is not only a criterion for an achievement of a good adhesion and wear

behaviour of the complex substrate – thin film. There is probably an optimal combination of

hardness and composition of the layer at which the adhesion is maximal.

The specimens differ also in the texture of the layer. The diffraction peaks Cr2N(110) at

37o of the two theta angle and Cr2N(300) at 67

o are higher for the specimen 72 with a poor

adhesion than for the specimen 71 with good adhesion. The differences indicate to a different

mechanism of growth of both layers. On first sight this is surprising, since the layers were

prepared on the specimens processed by the same way as in the heat treatment stage and in the

nitriding stage and it is also in disagreement with earlier published data [8,10-12] where no

texture in direction (300) was reported. On the other hand, in abovementioned investigations

other steels were used as substrates and thus no direct comparison of results is available.

Figs. 12 (left), 13 (right) – Depth profiles of elements throughout the near surface region,

specimen with a good adhesion (12) and worse adhesion (13)

Fig. 14 – X-ray diffraction patterns from PVD CrN – layers

Another possible explanation of different wear behaviour of the specimens is a possible

exceeding of the carrying capacity of the substrate. Nevertheless, the scratch test indicated a

great scattering of critical load values and therefore, the differences in the layer formation

Specimen 71

CrFe

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12

Distance from the surface (µm)

Ele

me

nt

co

nte

nt

(%)

C Cr Fe N

Specimen 72

Cr

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14

Distance from the surface (µµµµm)

Ele

me

nt

co

nte

nt

(%)

C Cr Fe N

Cr 2

N

Cr 2

N

(Cr)

Cr 2

N

Cr 2

N

Cr 2

N

(Cr)

Cr 2

N

Cr 2

N 72

71

Two theta diffraction angle


___________________________________________________________________________

9

displayed by X-ray diffraction measurements is more probable primary explanation of

different wear resistance of the specimens. The cause of different layer growth, in a given

batch, too, can be probably attributed to the positional effect of the specimen with respect to

the chromium source (target) during the sputtering. No other effects with a possible influence

on the layer growth are known up to now.

4. CONCLUSIONS

1) For all the combinations of plasma nitriding and CrN-coating, the critical load measured

with the scratch test was at least 50 N, which is considered to be an acceptable for industrial

applications. In some cases the adhesion was much better and the critical load exceeded 130

N, which can be considered as an excellent and unique value.

2) Wear testing at a load of 50 N did not show a significant scattering of results, the wear of

specimens proceeded in a steady manner and slightly better results were found for thinner

CrN-layers.

3) The use of a load of 150 N highlighted the difference in wear behaviour of the specimens,

also of those prepared in the same processing conditions. For the specimens with an excellent

adhesion, the wear rate on the specimen´s side was low and on the counterpart´s side much

greater. The friction coefficient increased slightly with the increased sliding distance. For the

specimens with poor adhesion, a sudden and dramatical increase of wear rate on the

specimen´s side occurred. This is connected with the decrease of the friction coefficient and

the lowering of the wear on the counterpart´s side.

4) In case of specimens with excellent adhesion the CrN-coating remained on the surface,

although in a slightly damaged form. Coatings with small adherence cracked and were

progressively removed from the contact surface.

5) Layers with excellent adhesion differ from those with poor adhesion in chemistry and

phase constitution. In the firsts less chromium in solid solution and more Cr2N-compound

than the seconds were found. Moreover, the different texture indicates probably also to a

difference in growth mechanism.

6) The main challenge into the future is then: optimization of pre-nitriding in order to

minimize the loss in fracture toughness and investigation of the way how to lower the scatter

of adhesion values and wear resistance.

ACKNOWLEDGEMENTS

Authors wish to thank the Ministry of Education and Youth of the Czech Republic and

the Slovak Republic for the financial support for the solution of the Project Eureka E!3437

PROSURFMET

REFERENCES

[1]: Jurči, P., Panjan, P.: Surface Processing of the PM Vanadis 6 Steel with Plasma Nitriding

and CrN PVD – Coating, In: Proceedings of the European Powder Metallurgy Congress,

Prague, Czech Republic, 2005.

[2]: Jurči, P., Hnilica, F., Suchánek, J., Stolař, P.: Materiali in Tehnologije, 38, 2004, 1-2, 13.

[3]: Aubert, R., Gillet, A., Gaucher, J., Errat, P.: Thin Solid Films 108 (1983) 165.

[4]: Mayrhofer, P.H., Willmann, H., Mitterer, C.: Surf. Coat. Techn. 146-147 (2001) 222.

[5]: Lousa, A., Romero, J., Martinez, E., Esteve, J., Montala, F., Carreras, L.: Surf. Coat.

Techn. 146-147 (2001) 268.

[6]: Kondo, A., Oogami, T., Sato, K., Tanaka, Y., Surf. Coat. Techn. 177-178 (2004) 238.

[7]: Nouveau, C., Jorand, E., Deces-Petit, C., Labidi, C, Djouadi, M.A.: Wear 258 (2005) 157.

[8]: Odén, M., Almer, J., Hakansson, G., Olsson, M.: Thin Solid Films 377-378 (2000) 407.


___________________________________________________________________________

10

[9]: Lamastra F.R., Leonardi, F., Montanari, R., Casadei, F., Valente, T., Gusmano, G.: Surf.

Coat. Techn. 200 (2006) 6172.

[10]: Attar, F., Johannesson, T.: Thin Solid Films 258 (1995) 205.

[11]: Aharonov, R.R., Coll, B.F., Fontana, R.P.: Surf. Coat. Techn. 61 (1993) 223.

[12]: Grant, W.K., Loomis, C., Moore, J.J., Olson, D.L., Mishra, B., Perry, A.J.: Surf. Coat.

Techn. 86 – 87 (1996) 788.

[13]: Cunha, L., Andritschky, M., Pischow, K., Wang, Z.: Thin Solid Films 355 – 356 (1999)

465.

[14]: Broszeit, E., Friedrich, C., Berg, G.: Surf. Coat. Techn. 115 (1999) 9.

[15]: Mayrhofer, P.H., Tischler, G., Mitterer, C.: Surf. Coat. Techn. 142-144 (2001) 78.

[16]: Mercs, D., Bonasso, N., Naamane, S., Bordes, J.M., Coddet, C.: Surf. Coat. Techn. 200

(2005) 403.

[17]: Djouadi, M.A., Nouveau, C., Beer, P., Lambertin, M.: Surf. Coat. Techn. 133-134

(2000) 478.

[18]: Schell, N., Petersen, J.H., Bottiger, J., Mucklich, A., Chevallier, J., Andreasen, K.P.,

Eichhorn, F.: Thin Solid Films 426 (2003) 100.

[19]: Jurči, P., Musilová, A., Stolař, P., Hrubý, V.: In: Proceedings of the Conference

“Carburizing and Nitriding”, November 27-28, 2001, Brno, Czech Republic, p. 81.

[20]: Jurči, P., Hudáková, M.: In: Proceedings of the 16th

Int. Conf. Metal 2007, May 2007,

Hradec n. Moravicí, Czech Republic, CD – ROM.

[21]: Diao, D.F. et al.: Tribology Intern., 27, 1994, 4, pp. 267-272.

Documents

FRACTOGRAPHICAL ANALYSIS OF WORN …metal2013.tanger.cz/files/proceedings/metal_08/Lists/...METAL 2008 13. –15. 5. 2008, Hradec nad Moravicí _____ 1 FRACTOGRAPHICAL ANALYSIS OF