- Email: [email protected]
Friction transition diagram considering the effects of oxide layer formed on wear track of AISI 1045 steel disk against TiN coated AISI 52100 steel ball in sliding
Surface and Coatings Technology 179 (2004) 1–9
Friction transition diagram considering the effects of oxide layer formed on wear track of AISI 1045 steel disk against TiN coated AISI 52100 steel ball in sliding Chung-Woo Choa, Young-Ze Leeb,* a
Graduate School of Mechanical Engineering, SungKyunKwan University, Suwon, KyungGi-Do 440-746, South Korea b School of Mechanical Engineering, SungKyunKwan University, Suwon, KyungGi-Do 440-746, South Korea Received 22 October 2001; accepted in revised form 23 May 2003
Abstract In this study, the effect of the real contact area on oxide layer formation was investigated by varying coating thickness, surface roughness of counter part and contact load between two materials in order to understand the wear mechanism of a TiN coated ball against a steel disk. After conducting sliding tests under different conditions, the characteristics of the oxide layer that are formed on the sliding surfaces of the two materials and changes in friction signal, which were caused by the oxide layer formation, were examined. For the coated ball specimen, an AISI 52100 steel ball was used and AISI 1045 steel was used for the disk counter part. From the results, the iron oxide layer on the steel disk dominates the friction characteristic between the two materials and induces a friction transition and high friction. The oxide layer, on the sliding surfaces of two materials, begins to form at a point through which friction signal transits from low friction to high friction. Also, the formation on the counter part occurred early as the contact load increased, as the surface roughness of counter part decreased and as the TiN coating thickness decreased. 䊚 2003 Elsevier B.V. All rights reserved. Keywords: TiN coated ball; Oxide layer; Real contact area; Adhesive wear; Friction transition diagram
1. Introduction Ceramic coatings can increase the life of machine elements due to their outstanding low friction characteristics and good wear resistance w1x. Therefore, ceramic coatings such as titanium nitride (TiN), chromium nitride (CrN) and titanium aluminum nitride (TiAlN), are generally being applied to many machine elements including machine tools, bearings and shafts w1,2x. A problem that arises when such coatings are applied to a part which is counter acting with other material, is that friction and wear characteristics of the coating change according to characteristics of the transfer layer, which forms, for example, due to contact with steel and an oxide layer, and due to transfer layer w3x. Many research results have reported that an oxide layer can protect coating material from wear when in contact with a *Corresponding author. ChangAn-Gu, chunchun-Do 300, 440.746 South Korea. Tel.: q82-31-290-7444; fax: q82-31-290-5276. E-mail address: [email protected] (Y.-Z. Lee).
counter part w4–7x. When a nitride ceramic coating is in contact with steel, it usually shows outstanding characteristics of low friction and wear resistance initially, but as sliding motion continues, a transfer layer and an iron oxide layer forms on sliding surfaces of both the coating and steel due to wear particles from steel. In this case, it is known that the coating is protected from wear once the layers are formed w4,6,7x, but the low friction characteristic of the coating may disappear and the friction force between two materials increases causing wear of the steel to increase, as contact condition changes from steel on coating to steel on oxide layer, oxide layer on coating or oxide layer on oxide layer. In this study, the effect of real contact area on oxide layer was investigated by varying coating thickness, surface roughness of counter part and contact load between two materials as a primary research for understanding the wear mechanism of TiN coating. After conducting sliding tests under different conditions, char-
0257-8972/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0257-8972(03)00801-6
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
2
cell. The signal from the load cell was stored in a computer at a sampling rate of 5 Hz after converting it using analogydigital converter. And then the stored signal was converted to coefficient of friction (COF) signal by signal processing program. 2.2. Materials and test conditions
Fig. 1. Schematic diagram of sliding tester: (a) Uncoated ball on steel disk with Ra 0.2 mm, (b) TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.2 mm.
acteristics of the oxide layers that are formed on the sliding surfaces of two materials and changes in friction signal, which were caused by the oxide layer formation, were examined. Based on the obtained results, contact load was increased while testing against various specimens and then the changes in friction signal between TiN coating and oxide layer were observed. As a result, the friction transition diagram was generated. 2. Experimental details 2.1. Sliding tester A generic ball-on-disk sliding tester was used as shown schematically in Fig. 1. The contact load was determined by applying deadweights on the coated ball directly. The friction force was measured using a load
For the coated ball specimen, an AISI 52100 steel ball was used and its diameter was 10 mm. Two types of balls were prepared by depositing TiN coating with 1 and 4 mm in coating thickness using arc ion plating method. AISI 1045 steel was used for the disk counter part with a diameter of 60 mm and thickness of 7 mm. The surface hardness was HV1N300. Three types of disks were prepared based on the surface roughness: Ra 0.06, 0.1 and 0.2 mm. Various TiN coating thicknesses and surface roughnesses of the steel disks were used to investigate the characteristics of oxide layer according to the changes in real contact area between two materials, due to the changes in coating thickness and surface roughness. A slow sliding speed of 0.04 mys (30 rev.ymin) was applied in all tests and the applied sliding contact load ranged from 0.3 to 0.6 N since it was found that the oxide layer forms very rapidly at high sliding speed and contact load throughout the pretests and related researches w3x, the test condition was determined as stated earlier to allow gradual formation of oxide layer on the contact parts. Also, to study the effect of oxide layer on friction characteristic, tests were carried out both in air and nitrogen environments. All specimens were ultracleaned with acetone before each test.
Fig. 2. COF signal from the sliding tests under 0.3 N of contact load in air: (a) In air, (b) In nitrogen.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
3
load were investigated by analyzing the friction signals. Also, the friction transition due to oxide layer formation was represented in a friction transition diagram. 2.3. X-ray diffraction (XRD) analysis X-ray diffraction (XRD) analysis was carried out on the wear tracks of the uncoated steel disk tested in air and nitrogen, using Bruker AXS D8 Discover X-ray diffractometer with Cu-Ka radiation. The diffraction patterns were acquired in a two-theta angle range of 20–808 at a scan speed of 4 8ymin. 3. Results and discussion 3.1. Friction characteristic of TiN coated ball and steel disk under sliding test in air Fig. 3. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.06 mm in various environments under 0.5 N of contact load. (a) In air, (b) In nitrogen.
During the tests, the friction signals under each condition were measured and then, based on these data, the effect of oxide layer on friction characteristics was investigated. Also, the characteristic of oxide layer formation according to the friction signal change was studied as well. As a result, the instance of oxide layer formation was defined through the friction signal analysis. Differences in the effect of real contact area between two materials on the formation of oxide layer due to the variations in ceramic coating thickness, surface roughness of counter part and sliding contact
Fig. 4. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.1 mm in various environments under 0.5 N of contact load. (a) In air, (b) In nitrogen.
To investigate the friction characteristic of the TiN coating, the changes in coefficient of friction were measured by putting both the coated ball and the uncoated ball in sliding motion against the steel disk. As shown in Fig. 2a, typical dry friction behaviors, the sudden transition and the high friction were observed as the contact number of cycle increased for the uncoated ball. For the coated ball, the friction signal showed a gradual transition behavior according to the contact number of cycles as shown in Fig. 2b. Such friction signal can be divided into three regions; the low friction region where typical friction characteristic of the TiN coating belongs, the friction transition region and the high friction region where the TiN coating characteristic does not exist in friction point of view. Although the
Fig. 5. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.2 mm in various environments under 0.5 N of contact load.
4
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
Fig. 6. The SEM micrographs and EDX line profiles of the wear tracks on steel disk generated during the sliding tests, which results are presented in Fig. 3: (a) Wear track generated in air, (b) wear track generated in nitrogen.
period of low friction region is constant when tested using same counter part under same conditions, it varies depending on the counterpart material. Most of the related research results reported that the transition from low friction to high friction, which appears during sliding test of ceramic coated ball and steel, was due to the damage of the ceramic coating w7– 12x. However, in sliding motion of a TiN coated ball against steel disk, a different behavior, which is the induction of high friction due to oxide layer formation without coating being damaged, can be observed. To verify such behavior, the following sliding tests were conducted under various oxide layer formation conditions. 3.2. Effect of oxide layer on friction characteristics To investigate the effect of oxide layer, which forms on the sliding surfaces of TiN coated ball and uncoated steel disk, on friction characteristic between the two
materials, sliding tests were conducted both in air and in nitrogen environments under the contact load of 0.5 N using TiN coated ball with 1 mm coating thickness and steel disks with 0.06, 0.1 and 0.2 mm. The results are represented in Figs. 3–5 for comparison. Fig. 3b, Fig. 4b and Fig. 5b are the test results from the test in nitrogen and it can be verified that the low friction region in ceramic coating continues during the tests unlike in air, which are shown in Fig. 3a, Fig. 4a and Fig. 5a. This proves that both the friction transition region and the high friction region, which appear in air, are generated due to the oxidation of sliding surfaces. Also, in Fig. 3b, Fig. 4b and Fig. 5b, the effect of surface roughness of steel disk on the friction characteristic of TiN coated ball, which is the decrease in coefficient of friction as the surface roughness increases, can be observed. This is because the real contact area between two materials reduces as the surface of counter part gets rougher w13x. Therefore, the wear and the friction between the two materials under the surface
Fig. 7. XRD spectra of the wear tracks on steel disk tested in air; %–Fe2O3.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
5
Fig. 8. COF signal, SEM micrographs and EDX analysis of the wear tracks on steel disks from the sliding tests of TiN coating ball with 1 mm coating thickness on disks with different surface roughness under 50 cycles and 0.3 N of contact load.
roughness range were effected more by adhesive wear due to the difference in real contact area rather than by abrasive wear due to asperity interlocking. Thus, adhesive wear is believed to be the major wear mechanism that occurs between the two.
Fig. 6a and b are the scanning electron microscopy (SEM) micrographs and the line profiles from the energy-dispersive X-ray spectroscopy (EDX) analysis of the wear tracks generated on steel disk from Fig. 3a and b, respectively. It can be seen that iron, the main
Fig. 9. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.2 mm under various contact loads in air.
Fig. 10. COF signal from the sliding tests of TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.2 mm under various contact loads in air.
6
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
element in steel disk, is distributed along the line profile, line AB, while titanium, the main element in TiN coating, distribution is at a noise level in Fig. 6a and b. This proves that material transfer from the coated ball to the steel disk did not occur. As shown in Fig. 6a, massive amount of oxygen was detected from the wear tracks generated during the tests in air. Also, the iron content at the peaks in the line profile of oxygen decreased due to the bonding of oxygen and iron. To investigate the material structures of oxide layers formed on the wear track generated from the test in air, XRD analysis was carried out using Cu-Ka excitation. The lattice parameters are compared with the powder diffraction database. Fe2O3 were identified on the wear track of the steel disk as shown in Fig. 7, but any TiO2 were not on the wear track. Thus, it can be said that the friction characteristic between the two materials in air is decided by the iron oxide layer (Fe2O3) that forms on the sliding surfaces and the contact condition changes by the oxide layer inducing high friction rather than the intrinsic low friction characteristic of TiN.
Fig. 12. COF signal from the sliding tests of TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.1 mm under various contact loads in air.
3.3. Characteristic of oxide layer formation according to friction signal
Fig. 11. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.1 mm under various contact loads in air.
To investigate the characteristic of oxide layer formation according to the friction signal, sliding tests were conducted using the coated ball with 1 mm coating thickness on steel disks with surface roughness of 0.06, 0.1 and 0.2 mm under 0.3 N of contact load for 50 cycles in contact. The wear tracks generated on steel disks were examined using SEM and analyzed by EDX. The results are shown in Fig. 8. As shown in Fig. 8a,b and c, three different friction signals were obtained according to the surface roughness. In Fig. 8a, it can be seen that the high friction region was obtained when the steel disk with Ra 0.06 mm was used. After analyzing the wear track with EDX, massive amount of oxygen was detected. For the steel disk with Ra 0.1 mm, the friction transition region was obtained as shown in Fig. 8b and EDX analysis showed that the oxygen amount in wear track decreased as compared to Fig. 8a. Fig. 8c is the case, which the steel disk with Ra 0.2 mm was used and the coefficient of friction stayed in the low friction region during the test.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
Fig. 13. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.06 mm under various contact loads in air.
Unlike in Fig. 8a and b, oxygen was not detected, which indicates no existence of oxide layer on the wear track. Thus, it can be said that the main wear mechanism, which forms oxide layer during sliding test of coated ball on steel disk is adhesive wear although the occurrence of abrasive wear due to the effect of surface roughness was verified. From the above test results, it was found that the oxide layer forms at a point where the friction signal transits from the low friction region to the high friction region. Also, the behavior of oxide layer formation changes according to the surface roughness of counter part. As the surface roughness increases, the oxide layer begins to form late and therefore, the friction transition occurs late. This indicates that, by observing the friction signal during sliding test, the initiation point of oxide layer formation can be predicted. Thus, the oxide layer forming behavior according to the real contact area can be investigated by observing the transition of friction signal as discussed in the following Section 3.4.
7
and the contact load. The initiation time of oxide layer formation was measured by counting the contact number of cycles up to the end of low friction region. In Figs. 9 and 10, the results of sliding tests, which slid the coated ball with 1 and 4 mm coating thickness on steel disk with Ra 0.2 mm, are shown, respectively. The applied contact loads were 0.3, 0.4, 0.5 and 0.6 N. As shown in the figures, the low friction region was decreased since the oxide layer formed early, as the contact load increased. Also, under the same contact load, the low friction region was increased in the coated ball with 4 mm coating thickness compared to the ball with 1 mm coating thickness since the real contact area was smaller due to the higher surface hardness of thicker coating ball. In other words, when real contact area is large, adhesive wear occurs easily causing oxide layer to form early and decreasing the low friction region. This behavior was observed also in the sliding tests using steel disks with Ra 0.1 and 0.06 mm. The results of sliding tests, which used the coated ball with 1 mm coating thickness and steel disks with Ra 0.1 and 0.06 mm, are presented in Figs. 11 and 13, respectively, and the ball with 4 mm coating thickness and the disks with Ra 0.1 and 0.06 mm in Figs. 12 and 14, respectively. As shown in Figs. 9 and 10, the decrease of low friction region was observed in Figs. 11–14 since the oxide layer formed early due to the easy occurrence of adhesive wear by the increase in real
3.4. Initiation time of oxide layer formation according to the real contact area of sliding surfaces To study the effect of real contact area between the two materials on the formation of the iron oxide layer, further sliding tests were conducted by varying the coating thickness, the surface roughness of steel disk
Fig. 14. COF signal from the sliding tests of TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.06 mm under various contact loads in air.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
8
Table 1 Contact number of cycle at oxide layer formation Surface roughness of the steel disks Ra 0.2 (mm)
Ra 0.1 (mm)
Ra 0.06 (mm)
Coating thickness
1 (mm)
4 (mm)
1 (mm)
4 (mm)
1 (mm)
4 (mm)
Normal load (N) 0.3 0.4 0.5 0.6
73 40 34 26
98 60 40 36
23 12 9 5
75 29 20 12
0 0 0 0
0 0 0 0
contact area as the contact load increased. In Figs. 9– 14, the characteristic of oxide layer formation according to the surface roughness of steel disk can be seen. Under the same contact load and the coating thickness, the low friction region increases as the surface roughness increases due to the late formation of oxide layer. Table 1 contains the list of contact number of cycle at the beginning of oxide layer formation according to the contact load, the coating thickness and the surface roughness of steel disks obtained from Figs. 9–14. By plotting these data in log–log scale, the friction transition diagram was generated as shown in Fig. 15. From the diagram, it can be seen that as the contact load increases, the contact number of cycle at the beginning of oxide layer formation decreases linearly. Also, it can be observed that the contact number of cycle at the beginning of oxide layer formation increases as the coating thickness increases and the surface roughness of steel disk increases under same contact load in the order of 1 mm-Ra 0.1 mm, 4 mm-Ra 0.1 mm, 1 mm-Ra 0.2 mm and 4 mm-Ra 0.2 mm. The decrease of real contact area is believed to be in the same order as well. In this
study, the effect of surface roughness of steel disk on the real contact area size is found to be more significant than the effect of coating thickness. 4. Conclusions By varying the coating thickness and the surface roughness of counter part, the effect of real contact area on the oxide layer formation and the wear mechanism of TiN coated ball was investigated and as a result, the following conclusions were obtained. 1. The iron oxide layer on steel disk dominates the friction characteristic between the two materials and it induces friction transition and high friction. 2. Within the surface roughness range used in this study, the friction between two materials was more affected by the real contact area rather than the asperity interlocking. Also, when the surface of counter part was rough, the initial wear mechanism in the coated ball was an abrasive wear and oxide layer did not form by it. 3. The oxide layer, on the sliding surfaces of two materials, begins to form at a point where friction signal transits from low friction to high friction. The formation on the counter part occurred early as the contact load increased, as the surface roughness of counter part decreased and as TiN coating thickness decreased. That is, it forms early as the real contact area increases. 4. The main wear mechanism for oxide layer formation is adhesive wear based on the conclusions (2) and (3). Acknowledgments The authors are grateful for the support provided by a grant from the Safety and Structural Integrity Research Center at SungKyunKwan University. References
Fig. 15. Friction transition diagram; contact number of cycle at oxide layer formation as a function of contact load.
w1x K. Holmberg, A. Matthews, Coat. Tribol. Elsevier (1994) 172–189.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9 w2x B. Bhushan, B.K. Gupta, Handbook of Tribology, McGrawHill, 1991, pp. 9.1–9.121. w3x K. Holmberg, H. Ronkainen, A. Matthews, Ceram. Int. 26 (2000) 787–795. w4x S. Wilson, A.T. Alpas, Surf. Coat. Technol. 94–95 (1997) 53–59. w5x S. Wilson, A.T. Alpas, Wear 245 (2000) 223–229. w6x Z.P. Huang, Y. Sun, T. Bell, Wear 173 (1994) 13–20. w7x A. Erdemir, C. Bindal, J. Pagan, P. Wilbur, Surf. Coat. Technol. 76–77 (1995) 559–563.
9
w8x D. Dowson, C.M. Taylor, T.H.C. Childs, M. Godet, G. Dalmaz, Thin films in tribology, Tribology series 25, Elsevier, 1993. w9x K. Holmberg, H. Ronkainen, A. Matthews, Wear mechanisms of coated sliding surfaces, Tribology series 25, Elsevier, 1993. w10x H. Sander, D. Petersohn, Friction and wear behavior of PVDcoated tribosystems, Tribology series 25, Elsevier, 1993. w11x C.-W. Cho, Y.-Z. Lee, Surf. Coat. Technol. 127 (2000) 59–65. w12x Y.-Z. Lee, K.-H. Jeong, Wear 217 (1998) 175–181. w13x E. Rabinowicz, Friction and Wear of Materials, Wiley, 1995, pp. 66–81.
Friction transition diagram considering the effects of oxide layer formed on wear track of AISI 1045 steel disk against TiN coated AISI 52100 steel ball in sliding Chung-Woo Choa, Young-Ze Leeb,* a
Graduate School of Mechanical Engineering, SungKyunKwan University, Suwon, KyungGi-Do 440-746, South Korea b School of Mechanical Engineering, SungKyunKwan University, Suwon, KyungGi-Do 440-746, South Korea Received 22 October 2001; accepted in revised form 23 May 2003
Abstract In this study, the effect of the real contact area on oxide layer formation was investigated by varying coating thickness, surface roughness of counter part and contact load between two materials in order to understand the wear mechanism of a TiN coated ball against a steel disk. After conducting sliding tests under different conditions, the characteristics of the oxide layer that are formed on the sliding surfaces of the two materials and changes in friction signal, which were caused by the oxide layer formation, were examined. For the coated ball specimen, an AISI 52100 steel ball was used and AISI 1045 steel was used for the disk counter part. From the results, the iron oxide layer on the steel disk dominates the friction characteristic between the two materials and induces a friction transition and high friction. The oxide layer, on the sliding surfaces of two materials, begins to form at a point through which friction signal transits from low friction to high friction. Also, the formation on the counter part occurred early as the contact load increased, as the surface roughness of counter part decreased and as the TiN coating thickness decreased. 䊚 2003 Elsevier B.V. All rights reserved. Keywords: TiN coated ball; Oxide layer; Real contact area; Adhesive wear; Friction transition diagram
1. Introduction Ceramic coatings can increase the life of machine elements due to their outstanding low friction characteristics and good wear resistance w1x. Therefore, ceramic coatings such as titanium nitride (TiN), chromium nitride (CrN) and titanium aluminum nitride (TiAlN), are generally being applied to many machine elements including machine tools, bearings and shafts w1,2x. A problem that arises when such coatings are applied to a part which is counter acting with other material, is that friction and wear characteristics of the coating change according to characteristics of the transfer layer, which forms, for example, due to contact with steel and an oxide layer, and due to transfer layer w3x. Many research results have reported that an oxide layer can protect coating material from wear when in contact with a *Corresponding author. ChangAn-Gu, chunchun-Do 300, 440.746 South Korea. Tel.: q82-31-290-7444; fax: q82-31-290-5276. E-mail address: [email protected] (Y.-Z. Lee).
counter part w4–7x. When a nitride ceramic coating is in contact with steel, it usually shows outstanding characteristics of low friction and wear resistance initially, but as sliding motion continues, a transfer layer and an iron oxide layer forms on sliding surfaces of both the coating and steel due to wear particles from steel. In this case, it is known that the coating is protected from wear once the layers are formed w4,6,7x, but the low friction characteristic of the coating may disappear and the friction force between two materials increases causing wear of the steel to increase, as contact condition changes from steel on coating to steel on oxide layer, oxide layer on coating or oxide layer on oxide layer. In this study, the effect of real contact area on oxide layer was investigated by varying coating thickness, surface roughness of counter part and contact load between two materials as a primary research for understanding the wear mechanism of TiN coating. After conducting sliding tests under different conditions, char-
0257-8972/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0257-8972(03)00801-6
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
2
cell. The signal from the load cell was stored in a computer at a sampling rate of 5 Hz after converting it using analogydigital converter. And then the stored signal was converted to coefficient of friction (COF) signal by signal processing program. 2.2. Materials and test conditions
Fig. 1. Schematic diagram of sliding tester: (a) Uncoated ball on steel disk with Ra 0.2 mm, (b) TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.2 mm.
acteristics of the oxide layers that are formed on the sliding surfaces of two materials and changes in friction signal, which were caused by the oxide layer formation, were examined. Based on the obtained results, contact load was increased while testing against various specimens and then the changes in friction signal between TiN coating and oxide layer were observed. As a result, the friction transition diagram was generated. 2. Experimental details 2.1. Sliding tester A generic ball-on-disk sliding tester was used as shown schematically in Fig. 1. The contact load was determined by applying deadweights on the coated ball directly. The friction force was measured using a load
For the coated ball specimen, an AISI 52100 steel ball was used and its diameter was 10 mm. Two types of balls were prepared by depositing TiN coating with 1 and 4 mm in coating thickness using arc ion plating method. AISI 1045 steel was used for the disk counter part with a diameter of 60 mm and thickness of 7 mm. The surface hardness was HV1N300. Three types of disks were prepared based on the surface roughness: Ra 0.06, 0.1 and 0.2 mm. Various TiN coating thicknesses and surface roughnesses of the steel disks were used to investigate the characteristics of oxide layer according to the changes in real contact area between two materials, due to the changes in coating thickness and surface roughness. A slow sliding speed of 0.04 mys (30 rev.ymin) was applied in all tests and the applied sliding contact load ranged from 0.3 to 0.6 N since it was found that the oxide layer forms very rapidly at high sliding speed and contact load throughout the pretests and related researches w3x, the test condition was determined as stated earlier to allow gradual formation of oxide layer on the contact parts. Also, to study the effect of oxide layer on friction characteristic, tests were carried out both in air and nitrogen environments. All specimens were ultracleaned with acetone before each test.
Fig. 2. COF signal from the sliding tests under 0.3 N of contact load in air: (a) In air, (b) In nitrogen.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
3
load were investigated by analyzing the friction signals. Also, the friction transition due to oxide layer formation was represented in a friction transition diagram. 2.3. X-ray diffraction (XRD) analysis X-ray diffraction (XRD) analysis was carried out on the wear tracks of the uncoated steel disk tested in air and nitrogen, using Bruker AXS D8 Discover X-ray diffractometer with Cu-Ka radiation. The diffraction patterns were acquired in a two-theta angle range of 20–808 at a scan speed of 4 8ymin. 3. Results and discussion 3.1. Friction characteristic of TiN coated ball and steel disk under sliding test in air Fig. 3. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.06 mm in various environments under 0.5 N of contact load. (a) In air, (b) In nitrogen.
During the tests, the friction signals under each condition were measured and then, based on these data, the effect of oxide layer on friction characteristics was investigated. Also, the characteristic of oxide layer formation according to the friction signal change was studied as well. As a result, the instance of oxide layer formation was defined through the friction signal analysis. Differences in the effect of real contact area between two materials on the formation of oxide layer due to the variations in ceramic coating thickness, surface roughness of counter part and sliding contact
Fig. 4. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.1 mm in various environments under 0.5 N of contact load. (a) In air, (b) In nitrogen.
To investigate the friction characteristic of the TiN coating, the changes in coefficient of friction were measured by putting both the coated ball and the uncoated ball in sliding motion against the steel disk. As shown in Fig. 2a, typical dry friction behaviors, the sudden transition and the high friction were observed as the contact number of cycle increased for the uncoated ball. For the coated ball, the friction signal showed a gradual transition behavior according to the contact number of cycles as shown in Fig. 2b. Such friction signal can be divided into three regions; the low friction region where typical friction characteristic of the TiN coating belongs, the friction transition region and the high friction region where the TiN coating characteristic does not exist in friction point of view. Although the
Fig. 5. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.2 mm in various environments under 0.5 N of contact load.
4
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
Fig. 6. The SEM micrographs and EDX line profiles of the wear tracks on steel disk generated during the sliding tests, which results are presented in Fig. 3: (a) Wear track generated in air, (b) wear track generated in nitrogen.
period of low friction region is constant when tested using same counter part under same conditions, it varies depending on the counterpart material. Most of the related research results reported that the transition from low friction to high friction, which appears during sliding test of ceramic coated ball and steel, was due to the damage of the ceramic coating w7– 12x. However, in sliding motion of a TiN coated ball against steel disk, a different behavior, which is the induction of high friction due to oxide layer formation without coating being damaged, can be observed. To verify such behavior, the following sliding tests were conducted under various oxide layer formation conditions. 3.2. Effect of oxide layer on friction characteristics To investigate the effect of oxide layer, which forms on the sliding surfaces of TiN coated ball and uncoated steel disk, on friction characteristic between the two
materials, sliding tests were conducted both in air and in nitrogen environments under the contact load of 0.5 N using TiN coated ball with 1 mm coating thickness and steel disks with 0.06, 0.1 and 0.2 mm. The results are represented in Figs. 3–5 for comparison. Fig. 3b, Fig. 4b and Fig. 5b are the test results from the test in nitrogen and it can be verified that the low friction region in ceramic coating continues during the tests unlike in air, which are shown in Fig. 3a, Fig. 4a and Fig. 5a. This proves that both the friction transition region and the high friction region, which appear in air, are generated due to the oxidation of sliding surfaces. Also, in Fig. 3b, Fig. 4b and Fig. 5b, the effect of surface roughness of steel disk on the friction characteristic of TiN coated ball, which is the decrease in coefficient of friction as the surface roughness increases, can be observed. This is because the real contact area between two materials reduces as the surface of counter part gets rougher w13x. Therefore, the wear and the friction between the two materials under the surface
Fig. 7. XRD spectra of the wear tracks on steel disk tested in air; %–Fe2O3.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
5
Fig. 8. COF signal, SEM micrographs and EDX analysis of the wear tracks on steel disks from the sliding tests of TiN coating ball with 1 mm coating thickness on disks with different surface roughness under 50 cycles and 0.3 N of contact load.
roughness range were effected more by adhesive wear due to the difference in real contact area rather than by abrasive wear due to asperity interlocking. Thus, adhesive wear is believed to be the major wear mechanism that occurs between the two.
Fig. 6a and b are the scanning electron microscopy (SEM) micrographs and the line profiles from the energy-dispersive X-ray spectroscopy (EDX) analysis of the wear tracks generated on steel disk from Fig. 3a and b, respectively. It can be seen that iron, the main
Fig. 9. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.2 mm under various contact loads in air.
Fig. 10. COF signal from the sliding tests of TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.2 mm under various contact loads in air.
6
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
element in steel disk, is distributed along the line profile, line AB, while titanium, the main element in TiN coating, distribution is at a noise level in Fig. 6a and b. This proves that material transfer from the coated ball to the steel disk did not occur. As shown in Fig. 6a, massive amount of oxygen was detected from the wear tracks generated during the tests in air. Also, the iron content at the peaks in the line profile of oxygen decreased due to the bonding of oxygen and iron. To investigate the material structures of oxide layers formed on the wear track generated from the test in air, XRD analysis was carried out using Cu-Ka excitation. The lattice parameters are compared with the powder diffraction database. Fe2O3 were identified on the wear track of the steel disk as shown in Fig. 7, but any TiO2 were not on the wear track. Thus, it can be said that the friction characteristic between the two materials in air is decided by the iron oxide layer (Fe2O3) that forms on the sliding surfaces and the contact condition changes by the oxide layer inducing high friction rather than the intrinsic low friction characteristic of TiN.
Fig. 12. COF signal from the sliding tests of TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.1 mm under various contact loads in air.
3.3. Characteristic of oxide layer formation according to friction signal
Fig. 11. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.1 mm under various contact loads in air.
To investigate the characteristic of oxide layer formation according to the friction signal, sliding tests were conducted using the coated ball with 1 mm coating thickness on steel disks with surface roughness of 0.06, 0.1 and 0.2 mm under 0.3 N of contact load for 50 cycles in contact. The wear tracks generated on steel disks were examined using SEM and analyzed by EDX. The results are shown in Fig. 8. As shown in Fig. 8a,b and c, three different friction signals were obtained according to the surface roughness. In Fig. 8a, it can be seen that the high friction region was obtained when the steel disk with Ra 0.06 mm was used. After analyzing the wear track with EDX, massive amount of oxygen was detected. For the steel disk with Ra 0.1 mm, the friction transition region was obtained as shown in Fig. 8b and EDX analysis showed that the oxygen amount in wear track decreased as compared to Fig. 8a. Fig. 8c is the case, which the steel disk with Ra 0.2 mm was used and the coefficient of friction stayed in the low friction region during the test.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
Fig. 13. COF signal from the sliding tests of TiN coating ball with 1 mm coating thickness on steel disk with Ra 0.06 mm under various contact loads in air.
Unlike in Fig. 8a and b, oxygen was not detected, which indicates no existence of oxide layer on the wear track. Thus, it can be said that the main wear mechanism, which forms oxide layer during sliding test of coated ball on steel disk is adhesive wear although the occurrence of abrasive wear due to the effect of surface roughness was verified. From the above test results, it was found that the oxide layer forms at a point where the friction signal transits from the low friction region to the high friction region. Also, the behavior of oxide layer formation changes according to the surface roughness of counter part. As the surface roughness increases, the oxide layer begins to form late and therefore, the friction transition occurs late. This indicates that, by observing the friction signal during sliding test, the initiation point of oxide layer formation can be predicted. Thus, the oxide layer forming behavior according to the real contact area can be investigated by observing the transition of friction signal as discussed in the following Section 3.4.
7
and the contact load. The initiation time of oxide layer formation was measured by counting the contact number of cycles up to the end of low friction region. In Figs. 9 and 10, the results of sliding tests, which slid the coated ball with 1 and 4 mm coating thickness on steel disk with Ra 0.2 mm, are shown, respectively. The applied contact loads were 0.3, 0.4, 0.5 and 0.6 N. As shown in the figures, the low friction region was decreased since the oxide layer formed early, as the contact load increased. Also, under the same contact load, the low friction region was increased in the coated ball with 4 mm coating thickness compared to the ball with 1 mm coating thickness since the real contact area was smaller due to the higher surface hardness of thicker coating ball. In other words, when real contact area is large, adhesive wear occurs easily causing oxide layer to form early and decreasing the low friction region. This behavior was observed also in the sliding tests using steel disks with Ra 0.1 and 0.06 mm. The results of sliding tests, which used the coated ball with 1 mm coating thickness and steel disks with Ra 0.1 and 0.06 mm, are presented in Figs. 11 and 13, respectively, and the ball with 4 mm coating thickness and the disks with Ra 0.1 and 0.06 mm in Figs. 12 and 14, respectively. As shown in Figs. 9 and 10, the decrease of low friction region was observed in Figs. 11–14 since the oxide layer formed early due to the easy occurrence of adhesive wear by the increase in real
3.4. Initiation time of oxide layer formation according to the real contact area of sliding surfaces To study the effect of real contact area between the two materials on the formation of the iron oxide layer, further sliding tests were conducted by varying the coating thickness, the surface roughness of steel disk
Fig. 14. COF signal from the sliding tests of TiN coating ball with 4 mm coating thickness on steel disk with Ra 0.06 mm under various contact loads in air.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9
8
Table 1 Contact number of cycle at oxide layer formation Surface roughness of the steel disks Ra 0.2 (mm)
Ra 0.1 (mm)
Ra 0.06 (mm)
Coating thickness
1 (mm)
4 (mm)
1 (mm)
4 (mm)
1 (mm)
4 (mm)
Normal load (N) 0.3 0.4 0.5 0.6
73 40 34 26
98 60 40 36
23 12 9 5
75 29 20 12
0 0 0 0
0 0 0 0
contact area as the contact load increased. In Figs. 9– 14, the characteristic of oxide layer formation according to the surface roughness of steel disk can be seen. Under the same contact load and the coating thickness, the low friction region increases as the surface roughness increases due to the late formation of oxide layer. Table 1 contains the list of contact number of cycle at the beginning of oxide layer formation according to the contact load, the coating thickness and the surface roughness of steel disks obtained from Figs. 9–14. By plotting these data in log–log scale, the friction transition diagram was generated as shown in Fig. 15. From the diagram, it can be seen that as the contact load increases, the contact number of cycle at the beginning of oxide layer formation decreases linearly. Also, it can be observed that the contact number of cycle at the beginning of oxide layer formation increases as the coating thickness increases and the surface roughness of steel disk increases under same contact load in the order of 1 mm-Ra 0.1 mm, 4 mm-Ra 0.1 mm, 1 mm-Ra 0.2 mm and 4 mm-Ra 0.2 mm. The decrease of real contact area is believed to be in the same order as well. In this
study, the effect of surface roughness of steel disk on the real contact area size is found to be more significant than the effect of coating thickness. 4. Conclusions By varying the coating thickness and the surface roughness of counter part, the effect of real contact area on the oxide layer formation and the wear mechanism of TiN coated ball was investigated and as a result, the following conclusions were obtained. 1. The iron oxide layer on steel disk dominates the friction characteristic between the two materials and it induces friction transition and high friction. 2. Within the surface roughness range used in this study, the friction between two materials was more affected by the real contact area rather than the asperity interlocking. Also, when the surface of counter part was rough, the initial wear mechanism in the coated ball was an abrasive wear and oxide layer did not form by it. 3. The oxide layer, on the sliding surfaces of two materials, begins to form at a point where friction signal transits from low friction to high friction. The formation on the counter part occurred early as the contact load increased, as the surface roughness of counter part decreased and as TiN coating thickness decreased. That is, it forms early as the real contact area increases. 4. The main wear mechanism for oxide layer formation is adhesive wear based on the conclusions (2) and (3). Acknowledgments The authors are grateful for the support provided by a grant from the Safety and Structural Integrity Research Center at SungKyunKwan University. References
Fig. 15. Friction transition diagram; contact number of cycle at oxide layer formation as a function of contact load.
w1x K. Holmberg, A. Matthews, Coat. Tribol. Elsevier (1994) 172–189.
C.-W. Cho, Y.-Z. Lee / Surface and Coatings Technology 179 (2004) 1–9 w2x B. Bhushan, B.K. Gupta, Handbook of Tribology, McGrawHill, 1991, pp. 9.1–9.121. w3x K. Holmberg, H. Ronkainen, A. Matthews, Ceram. Int. 26 (2000) 787–795. w4x S. Wilson, A.T. Alpas, Surf. Coat. Technol. 94–95 (1997) 53–59. w5x S. Wilson, A.T. Alpas, Wear 245 (2000) 223–229. w6x Z.P. Huang, Y. Sun, T. Bell, Wear 173 (1994) 13–20. w7x A. Erdemir, C. Bindal, J. Pagan, P. Wilbur, Surf. Coat. Technol. 76–77 (1995) 559–563.
9
w8x D. Dowson, C.M. Taylor, T.H.C. Childs, M. Godet, G. Dalmaz, Thin films in tribology, Tribology series 25, Elsevier, 1993. w9x K. Holmberg, H. Ronkainen, A. Matthews, Wear mechanisms of coated sliding surfaces, Tribology series 25, Elsevier, 1993. w10x H. Sander, D. Petersohn, Friction and wear behavior of PVDcoated tribosystems, Tribology series 25, Elsevier, 1993. w11x C.-W. Cho, Y.-Z. Lee, Surf. Coat. Technol. 127 (2000) 59–65. w12x Y.-Z. Lee, K.-H. Jeong, Wear 217 (1998) 175–181. w13x E. Rabinowicz, Friction and Wear of Materials, Wiley, 1995, pp. 66–81.