- Email: [email protected]
Effects of carbon content on wear property in pearlitic steels
Wear 253 (2002) 107–113
Effects of carbon content on wear property in pearlitic steels Masaharu Ueda a,∗ , Koichi Uchino a , Akira Kobayashi b a
Yawata Research and Development Laboratory, Nippon Steel Corporation, 1-1 Tobihata-cho, Tobata-ku, Kitakyushu 804-8501, Japan b Yawata Works, Nippon Steel Corporation, 1-1 Tobihata-cho, Tobata-ku, Kitakyushu 804-8501, Japan
Abstract To clarify the effects of carbon content on the rolling contact wear in steels, the authors conducted a two-cylinder rolling contact wear test using pearlitic steels with a carbon content in a range from 0.8 to 1.0 mass% and studied the relationship between the carbon content and the rolling contact wear. In addition, the authors examined the dominating factor in the rolling contact wear in pearlitic steels and the work-hardening rate of the rolling contact surface. The main findings obtained are as follows: (1) The wear resistance of pearlitic steels improve as carbon content increases. (2) The dominating factor in the rolling contact wear of pearlitic steels is the rolling contact surface hardness (RCSH). (3) The improved wear resistance of pearlitic steels is attributable to an increase in RCSH due to raising the work-hardening rate of the rolling contact surface as carbon content increases. (4) The reason why the work-hardening rate of the rolling contact surface of pearlitic steel rises as carbon content increases is considered to be as follows: an increase in the cementite density increases the amount of dislocation in the matrix ferrite and promotes the grain refinement of the matrix ferrite. As a result, the matrix ferrite is strengthened through the promotion of dislocation hardening and grain refinement. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Wear resistance; Rail; Pearlitic steel; Work-hardening; Rolling contact surface hardness (RCSH)
1. Introduction In recent years, railway companies have been increasing the speed of passenger trains and the load of freight trains to increase the efficiency of rail transport. This trend has been increasing the severity of the environment in which rails are used. Especially on heavy-haul railways in North America, the increasing wear of the railhead and the increasing frequency of internal fatigue defects in the railhead on curves have been causing a reduction in rail service life. Consequently, the railway companies have been calling for the development of rails having a longer life. A rail is worn by rolling contact involving slip with the wheel. The mechanism of rail wear is considered to be an adhesive wear, based on the minute shear fracture and plastic deformation of rolling contact surfaces [1,2]. Therefore, how to increase the hardness of rail steels as an effective measure for improving the wear resistance of rails has been studied. In conventional rail steels with a pearlite structure, various high-strength rails have been developed by decreasing the ∗ Corresponding author. Tel.: +81-93-872-6373; fax: +81-93-872-6809. E-mail address: [email protected] (M. Ueda).
lamellar spacing of the pearlite structure, thus contributing to a longer rail life on curves [3–5]. The rolling contact wear property of rail steels is correlated with the hardness, as confirmed in the case of pearlitic steels. On the other hand, it is reported that rolling contact wear property of rail steels can not be evaluated by their hardness alone because the amount of wear greatly varies depending on changes in the microstructure and the carbon content of steels [6–9]. However, the reason why rolling contact wear varies depending on these points have not been fully explained. To clarify the effects of carbon content on the rolling contact wear in steels, the authors conducted a two-cylinder rolling contact wear test using pearlitic steels with a carbon content in a range from 0.8 to 1.0 mass%, and studied the relationship between the carbon content and the rolling contact wear. In addition, based on a microhardness measurement and microstructure observation of rolling contact surfaces, the authors examined the dominating factor in the rolling contact wear in pearlitic steels and the effects of carbon content on this factor, and discussed the mechanism for the change in the microstructure of the rolling contact surface and work-hardening rate of the rolling contact surface.
0043-1648/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 3 - 1 6 4 8 ( 0 2 ) 0 0 0 8 9 - 3
108
M. Ueda et al. / Wear 253 (2002) 107–113
Table 1 Chemical composition ranges and hardness range of test steels. Chemical composition (mass%)
Hardness (Hv, 98 N)
C
Si
Mn
P
S
Cr
0.78–1.01
0.18–0.52
0.48–1.01
≤0.023
≤0.020
0.15–0.25
294–395
A 50 kg ingot (1250 ◦ C for 2 h) → hot rolling (40 t) → slack quench or salt-bath.
Finally, newly developed Hyper-Eutectoid steel rails are introduced.
2. Test methods 2.1. Test steels The chemical composition and hardness ranges of the test pearlitic steels are given in Table 1. The test steels had carbon levels of 0.8, 0.9 and 1.0 mass%, and were changed in hardness in the range from 294 to 395 Hv by heat treatment. These pearlitic steels are called 0.8, 0.9 and 1.0 mass%C steel, respectively in the following discussion. Each steel was produced by vacuum melting electrolytic iron and alloy, casting the molten steel into an ingot, and hot rolling ingot into a 40 mm thick plate. Before specimen preparation, each steel plate was reheated and either subject to accelerated-cooling or to salt-bath heat treatment to vary the hardness of the pearlite structure and to prevent the formation of a proeutectoid cementite structure. 2.2. Rolling contact wear test Fig. 1 shows the dimensions and shapes of rolling contact wear test specimens and schematically illustrates a rolling contact wear test machine. A cylinder, measuring
30 mm in diameter and 8 mm in thickness, was machined from each steel plate and was used as a test specimen. A cylinder of the same diameter and thickness as mentioned was machined from the 0.8 mass%C steel with a hardness of 380 Hv and was used as a wheel specimen. The rolling contact wear property of each test specimen was evaluated by a Nishihara wear test machine capable of simulating the rolling contact wear of the rail/wheel. The test conditions were set at a maximum Hertzian contact pressure of 640 MPa and a slip ratio of 20% to simulate the extreme wearing condition of the railhead side in curved tracks. The rolling contact surface of the specimen was cooled with compressed air to prevent the change of the microstructure by heating and to remove wear debris. The weight loss was determined to be the difference between the pre-test weight and the post-test weight. The number of rolling cycles in this wear test was based on 700,000 cycles of the test specimen. To evaluate the wear property of the test specimens under rolling cycles, some wear tests were ended after 100,000; 300,000; or 500,000 cycles. After the test, the hardness of the rolling contact surface was measured with a microhardness tester, and its thin-film microstructure was observed with a transmission electron microscope (TEM).
3. Test results 3.1. Effect of pre-test hardness on weight loss Fig. 2 shows the relationship between the pre-test hardness and the weight loss of the test specimens. There is almost a linear relationship between the pre-test hardness and the weight loss. Namely, the weight loss decreases as the pre-test hardness increases. Moreover, there is a close correlation between the weight loss and the carbon content. 3.2. Effect of carbon content on weight loss
Fig. 1. Dimensions and shapes of wear test specimens and schematic illustration of rolling wear test machine.
To clarify the effect of the carbon content on the weight loss, the relationship between the carbon content and the weight loss was examined. Fig. 3 shows the relationship between the carbon content and the weight loss of the test specimens in the hardness range of 385–395 Hv. There is almost a linear correlation between the carbon content and the weight loss. Namely, the weight loss of the test specimens decreases as the carbon content increases.
M. Ueda et al. / Wear 253 (2002) 107–113
109
Fig. 2. Relationship between pre-test hardness and weight loss of test specimens.
Fig. 4. Relationship between number of rolling cycles and weight loss of test specimens with pre-test hardness range from 385 to 395 Hv.
Thus, it can be seen that the rolling contact wear property of the pearlitic steels greatly depend not only on the pre-test hardness but also on the carbon content.
It has been reported that the rolling contact wear property of rail steels is closely correlated with the number of rolling cycles and the post-test rolling contact surface hardness (RCSH) [10,11]. Accordingly, the authors investigated the relationship between the number of rolling cycles and the weight loss, and the relationship between the number of rolling cycles and the post-test RCSH by using test specimens with a pre-test hardness in the range from 385 to 395 Hv. The relationship between the number of rolling cycles
in the range from 100,000 to 700,000 and the weight loss of the test specimens is shown in Fig. 4. The relationship between the number of rolling cycles in the same range as mentioned and the post-test RCSH of the test specimens is shown in Fig. 5. There is little difference in weight loss in all of the test specimens up to 100,000 cycles. As the number of rolling cycles exceed 100,000, the weight loss changes in accordance with the carbon content, and an increase in the weight loss of the test specimens with a high carbon content is suppressed. As the number of rolling cycles increase further, the test specimens with a high carbon content have less weight loss compared with the test specimens with a low carbon content. The post-test RCSH in all the test specimens increase as the number of rolling cycles increase up to 100,000. As the number of rolling cycles exceed 100,000, the post-test
Fig. 3. Relationship between carbon content and weight loss of test specimens with pre-test hardness range from 385 to 395 Hv.
Fig. 5. Relationship between number of rolling cycles and post-test rolling contact surface hardness of test specimens with pre-test hardness range from 385 to 395 Hv.
3.3. Effects of number of rolling cycles on weight loss and post-test rolling contact surface hardness (RCSH)
110
M. Ueda et al. / Wear 253 (2002) 107–113
Fig. 6. TEM images beneath rolling contact surface of test specimens and selected area diffraction patterns after 700,000 cycles: (a) bright-field image of 0.8 mass%C steel; (b) dark-field image of 0.8 mass%C steel; (c) bright-field image of 1.0 mass%C steel; (d) dark-field image of 1.0 mass%C steel.
RCSH changes in accordance with the carbon content, and an increase in the post-test RCSH of the test specimens with a high carbon content is promoted. As the number of rolling cycles increase further, the post-test RCSH of the test specimens with a high carbon content is higher than that of the test specimens with a low carbon content. When the number of rolling cycles is 700,000, the post-test RCSH increases to 670 Hv for the 0.8 mass%C steel, to 710 Hv for the 0.9 mass%C steel, and to 750 Hv for the 1.0 mass%C steel.
the small area. Upon considering the difference between the carbon content and the size of crystalline grains in the dark field images, it is noted that the size of crystalline grains of the test specimen with a high carbon content is smaller than that of the test specimen with a low carbon content. This means that the matrix ferrite in the rolling contact surface is refined due to the increased carbon content of the pearlitic steels.
3.4. Observation of microstructure beneath rolling contact surface
4. Discussions
Fig. 6 shows TEM images beneath the rolling contact surface of the test specimens of the 0.8 and 1.0 mass%C steel after 700,000 cycles and the diffraction patterns obtained from a selected area of 1 mm in diameter. In the bright field images, the fracturing in the cementite phase, dislocations, and some dislocation cells are observed in the matrix ferrite in each test specimen. In the diffraction patterns of the photographed regions, diffraction spots of the cementite phase and the ferrite phase are dispersed, and the diffraction spots of the ferrite phase are formed in rings in each test specimen. This means that there are a lot of crystalline grains which are thought to be subgrains or nanocrystalline grains within
These test results showed that the change in the weight loss is closely correlated with the RCSH and that the RCSH depends on the number of rolling cycles in the wear test and the carbon content of the test specimens. In this chapter, the authors quantify the relationship between the RCSH and weight loss, and clarify the dominating factor in the rolling contact wear in pearlitic steels and the effects of carbon content on this factor. Further, the authors discuss the mechanism whereby the microstructure beneath the rolling contact surface varies as carbon content increases and the work-hardening rate of the rolling contact surface from the wear test results and the observation results of rolling contact surfaces.
M. Ueda et al. / Wear 253 (2002) 107–113
Fig. 7. Relationship between average rolling contact surface hardness and wear rate of test specimens at 200,000 cycle intervals between 100,000 and 700,000 cycles.
4.1. Dominating factor in rolling contact wear in pearlitic steels and effects of carbon content on this factor To clarify the relationship between the RCSH and the weight loss, the authors examined the relationship between the average RCSH and the wear rate in a range of rolling cycles where the RCSH and the weight loss clearly differed in accordance with the carbon content. Fig. 7 shows the relationship between the average RCSH and the wear rate (W) of test specimens at 200,000 cycle intervals between 100,000 and 700,000 cycles. There is almost a linear relationship between the average RCSH and the wear rate, and the carbon content is not a factor therein. Namely, the weight loss of pearlitic steel decreases as the RCSH increases. To confirm the effect of the carbon content on the RCSH, the authors, noting the difference between the pre- and post-test RCSH, set the ratio of the increase in the post-test RCSH to the pre-test hardness. The authors referred to this as the work-hardening rate of the rolling contact surface shown by Eq. (1), and examined the relationship between the carbon content and the work-hardening rate of the rolling contact surface. Fig. 8 shows the relationship between the carbon content of the test specimens and the work-hardening rate of the rolling contact surface. There is a little change in the work-hardening rate of the rolling contact surface at 100,000 cycles, at which the RCSH does not change as carbon content increases. However, there is a close relationship between the carbon content and the work-hardening rate of the rolling contact surface at 700,000 cycles, at which the RCSH rises as carbon content increases. That is, the work-hardening rate of the rolling contact surface rises as the carbon content increases after a
111
Fig. 8. Relationship between carbon content and work-hardening rate of rolling contact surface of test specimens.
period of rolling contact. Work-hardening rate of rolling contact surface (%) (post-test RCSH − pre-test hardness) × 100 = (pre-test hardness)
(1)
where the hardness values are all expressed in Vickers hardness (Hv). From these results, it is concluded that the dominating factor in the rolling contact wear of pearlitic steels is the RCSH and that the increase in the RCSH depends on the rise in the work-hardening rate of the rolling contact surface as the carbon content increases. Namely, it is thought that the improved wear resistance of pearlitic steels is attributable to an increase in RCSH due to raising the work-hardening rate of the rolling contact surface as the carbon content increases. 4.2. Mechanism for change in microstructure beneath rolling contact surface and rise in work-hardening rate of rolling contact surface with increase in carbon content According to these test results and observations, the reason why the work-hardening rate of the rolling contact surface of pearlitic steels rises is considered to be as follows: The rolling contact introduces strain into the ferrite phase and at the same time produces minute fractures in the cementite phase beneath the rolling contact surface. Repeated rolling contact concentrates strain in the ferrite phase that is lower in hardness than the fractured cementite. The strain concentration forms many dislocations in the matrix ferrite, and promotes dislocation hardening and grain refinement in the matrix ferrite. As a result, the matrix ferrite is strengthened by dislocation hardening and grain refinement. Moreover, the mechanism whereby the work-hardening rate of the rolling contact surface rises as the carbon content of the pearlitic steels increases is explained as follows: The
112
M. Ueda et al. / Wear 253 (2002) 107–113
increase in the carbon content of the pearlitic steels raises the density of the hard cementite phase. The increase in the cementite density promotes the concentration of strain in the matrix ferrite, and accelerates the grain refinement in the matrix ferrite. As a result, the matrix ferrite is strengthened further through the promotion of dislocation hardening and grain refinement as the carbon content increases. Tarui et al. reported that when a high carbon wire is cold-drawn, the cementite phase in the pearlite structure is decomposed, and the carbon of the cementite phase is dissolved into the matrix ferrite phase, and the ferrite phase is strengthened by the solid solution of carbon [12]. The rolling contact surface is subject to a more severe processing condition than that of the wire drawing, and the cementite phase is presumed to be decomposed by the rolling contact. Therefore, the increased amount of carbon dissolving into the matrix ferrite as the cementite density increases is considered to be one of the factors responsible for the raised work-hardening rate of the rolling contact surface.
5. Status of trial rails installation Based on the results of the laboratory test described, it was confirmed that the wear resistance in pearlitic steels could be improved by increasing the carbon content of these steels. To confirm the properties of rails in actual use, therefore, pearlitic steel rails with a carbon content of 0.9 mass% and hardness of HB 370 were produced on a trial basis and installed in actual tracks. In this chapter, one example of the wear property of these pearlitic steel rails on heavy-haul tracks in North America is described. The new rails were tested in an actual track and compared with the conventional pearlitic steel rails with a carbon content of 0.8 mass% and hardness of HB 370 (called DHH370). Fig. 9 shows the relationship between the passing tonnage
Fig. 9. Relationship between passing tonnage and amount of head side wear of rails in a curved track.
and the amount of head side wear of these rails in a curved track having a 290 m radius. As for the new rail, the amount of head side wear was less than that of the DHH370 rail in every passing tonnage. It was confirmed that the wear resistance was improved by increasing the carbon content of the pearlitic steels in the actual track. Moreover, when the service life of the new rail was compared with the DHH370 rail in half inches in the head side wear, it was clear that the service life of the new rail had been improved by about 55%. Upon confirming the improvement of the service life in actual tracks, the authors developed the Hyper-Eutectoid steel rails (named HE370, HE400) with the carbon content of 0.9 mass% and hardness of HB 370 and HB 400. It is thought that these rails will contribute in the future to greatly improving the service life of rails in heavy-haul railways where excellent wear resistance is demanded.
6. Conclusions Pearlitic steels with different carbon contents were two-cylinder wear tested to study the effect of the carbon content on the rolling contact wear. The dominating factor in the rolling contact wear in pearlitic steels and the effects of carbon content on this factor were examined. Further, the mechanism for the change in the microstructure beneath the rolling contact surface and the rise in the work-hardening rate of the rolling contact surface as the carbon content increases were discussed. The main findings obtained are as follows: (1) The wear resistance of pearlitic steels improve as the carbon content increases. (2) The dominating factor in the rolling contact wear of pearlitic steels is the RCSH. That is, the wear resistance of pearlitic steels improves as the RCSH increases. (3) The increase in the RCSH depends on the rise in the work-hardening rate of the rolling contact surface as the carbon content increases. (4) The improving wear resistance of pearlitic steels is attributable to an increase in the RCSH due to raising the work-hardening rate of the rolling contact surface as the carbon content increases. (5) The reason why the work-hardening rate of the rolling contact surface of pearlitic steel rises as the carbon content increases is considered to be as follows: an increase in the cementite density increases the amount of dislocation in the matrix ferrite and promotes the grain refinement of the matrix ferrite. As a result, the matrix ferrite is strengthened through the promotion of dislocation hardening and grain refinement. (6) Newly developed Hyper-Eutectoid steel rails (HE370) with a carbon content of 0.9 mass% show an improvement in service life of about 55% as compared with the conventional pearlitic steel rail (DHH370) with a carbon content of 0.8 mass% in an actual track.
M. Ueda et al. / Wear 253 (2002) 107–113
Acknowledgements The authors wish to thank Union Pacific Railroad and Burlington Northern Santa Fe for permitting the actual track tests to be conducted and Dr. Senuma for the fruitful discussions with him and his helpful comments. References [1] G.A. Tomlinson, Philos. Mag. 7 (1927) 905–911. [2] M. Ishida, Shinsenro 49 (1995) 38–40. [3] H. Masumoto, K. Sugino, H. Hayashida, in: Proceedings Conference on First International Heavy Haul Railway, Session 212, Colorado Springs, 1978. [4] K. Sugino, H. Kageyama, H. Masumoto, in: Proceedings Conference on Second International Heavy Haul Railway, Perth, 1982, pp. 187–197.
of the IHHA, of the IHHA,
113
[5] K. Sugino, H. Kageyama, C. Urashima, in: Proceedings of the 32nd Mechanical Working and Steel, ISS, Warrendale, 1990, 171–176. [6] P. Clayton, N. Jin, Wear 200 (1996) 74–82. [7] H. Yokoyama, S. Yamamoto, M. Fujikake, Y. Yoshida, in: Proceedings of the Fourth International Symposium on Rail Steels, ISS, Warrendale, 1997, pp. 1023–1028. [8] M. Ueda, K. Uchino, H. Kageyama, K. Kutaragi, K. Babazono, in: Proceedings of the Conference on Sixth International Heavy Haul Railway, IHHA, Cape Town, 1997, pp. 355–369. [9] M. Ueda, K. Uchino, H. Kageyama, M. Kaneta, A. Kobayashi, in: Proceedings of the Conference on IHHA’99 STS, Vol. 1, IHHA, Moscow, 1999, pp. 259–266. [10] K. Sugino, H. Masumoto, S. Nishida, C. Urashima, H. Kageyama, M. Hattori, Seitetsukenkyu 303 (1980) 23–38. [11] W.R. Tyfour, J.H. Beynon, A. Kupoor, Wear 180 (1995) 79– 89. [12] K. Tarui, N. Maruyama, G. Shigesato, H. Tashiro, Zairyou-to-Process 11 (1998) 1327.
Effects of carbon content on wear property in pearlitic steels Masaharu Ueda a,∗ , Koichi Uchino a , Akira Kobayashi b a
Yawata Research and Development Laboratory, Nippon Steel Corporation, 1-1 Tobihata-cho, Tobata-ku, Kitakyushu 804-8501, Japan b Yawata Works, Nippon Steel Corporation, 1-1 Tobihata-cho, Tobata-ku, Kitakyushu 804-8501, Japan
Abstract To clarify the effects of carbon content on the rolling contact wear in steels, the authors conducted a two-cylinder rolling contact wear test using pearlitic steels with a carbon content in a range from 0.8 to 1.0 mass% and studied the relationship between the carbon content and the rolling contact wear. In addition, the authors examined the dominating factor in the rolling contact wear in pearlitic steels and the work-hardening rate of the rolling contact surface. The main findings obtained are as follows: (1) The wear resistance of pearlitic steels improve as carbon content increases. (2) The dominating factor in the rolling contact wear of pearlitic steels is the rolling contact surface hardness (RCSH). (3) The improved wear resistance of pearlitic steels is attributable to an increase in RCSH due to raising the work-hardening rate of the rolling contact surface as carbon content increases. (4) The reason why the work-hardening rate of the rolling contact surface of pearlitic steel rises as carbon content increases is considered to be as follows: an increase in the cementite density increases the amount of dislocation in the matrix ferrite and promotes the grain refinement of the matrix ferrite. As a result, the matrix ferrite is strengthened through the promotion of dislocation hardening and grain refinement. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Wear resistance; Rail; Pearlitic steel; Work-hardening; Rolling contact surface hardness (RCSH)
1. Introduction In recent years, railway companies have been increasing the speed of passenger trains and the load of freight trains to increase the efficiency of rail transport. This trend has been increasing the severity of the environment in which rails are used. Especially on heavy-haul railways in North America, the increasing wear of the railhead and the increasing frequency of internal fatigue defects in the railhead on curves have been causing a reduction in rail service life. Consequently, the railway companies have been calling for the development of rails having a longer life. A rail is worn by rolling contact involving slip with the wheel. The mechanism of rail wear is considered to be an adhesive wear, based on the minute shear fracture and plastic deformation of rolling contact surfaces [1,2]. Therefore, how to increase the hardness of rail steels as an effective measure for improving the wear resistance of rails has been studied. In conventional rail steels with a pearlite structure, various high-strength rails have been developed by decreasing the ∗ Corresponding author. Tel.: +81-93-872-6373; fax: +81-93-872-6809. E-mail address: [email protected] (M. Ueda).
lamellar spacing of the pearlite structure, thus contributing to a longer rail life on curves [3–5]. The rolling contact wear property of rail steels is correlated with the hardness, as confirmed in the case of pearlitic steels. On the other hand, it is reported that rolling contact wear property of rail steels can not be evaluated by their hardness alone because the amount of wear greatly varies depending on changes in the microstructure and the carbon content of steels [6–9]. However, the reason why rolling contact wear varies depending on these points have not been fully explained. To clarify the effects of carbon content on the rolling contact wear in steels, the authors conducted a two-cylinder rolling contact wear test using pearlitic steels with a carbon content in a range from 0.8 to 1.0 mass%, and studied the relationship between the carbon content and the rolling contact wear. In addition, based on a microhardness measurement and microstructure observation of rolling contact surfaces, the authors examined the dominating factor in the rolling contact wear in pearlitic steels and the effects of carbon content on this factor, and discussed the mechanism for the change in the microstructure of the rolling contact surface and work-hardening rate of the rolling contact surface.
0043-1648/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 3 - 1 6 4 8 ( 0 2 ) 0 0 0 8 9 - 3
108
M. Ueda et al. / Wear 253 (2002) 107–113
Table 1 Chemical composition ranges and hardness range of test steels. Chemical composition (mass%)
Hardness (Hv, 98 N)
C
Si
Mn
P
S
Cr
0.78–1.01
0.18–0.52
0.48–1.01
≤0.023
≤0.020
0.15–0.25
294–395
A 50 kg ingot (1250 ◦ C for 2 h) → hot rolling (40 t) → slack quench or salt-bath.
Finally, newly developed Hyper-Eutectoid steel rails are introduced.
2. Test methods 2.1. Test steels The chemical composition and hardness ranges of the test pearlitic steels are given in Table 1. The test steels had carbon levels of 0.8, 0.9 and 1.0 mass%, and were changed in hardness in the range from 294 to 395 Hv by heat treatment. These pearlitic steels are called 0.8, 0.9 and 1.0 mass%C steel, respectively in the following discussion. Each steel was produced by vacuum melting electrolytic iron and alloy, casting the molten steel into an ingot, and hot rolling ingot into a 40 mm thick plate. Before specimen preparation, each steel plate was reheated and either subject to accelerated-cooling or to salt-bath heat treatment to vary the hardness of the pearlite structure and to prevent the formation of a proeutectoid cementite structure. 2.2. Rolling contact wear test Fig. 1 shows the dimensions and shapes of rolling contact wear test specimens and schematically illustrates a rolling contact wear test machine. A cylinder, measuring
30 mm in diameter and 8 mm in thickness, was machined from each steel plate and was used as a test specimen. A cylinder of the same diameter and thickness as mentioned was machined from the 0.8 mass%C steel with a hardness of 380 Hv and was used as a wheel specimen. The rolling contact wear property of each test specimen was evaluated by a Nishihara wear test machine capable of simulating the rolling contact wear of the rail/wheel. The test conditions were set at a maximum Hertzian contact pressure of 640 MPa and a slip ratio of 20% to simulate the extreme wearing condition of the railhead side in curved tracks. The rolling contact surface of the specimen was cooled with compressed air to prevent the change of the microstructure by heating and to remove wear debris. The weight loss was determined to be the difference between the pre-test weight and the post-test weight. The number of rolling cycles in this wear test was based on 700,000 cycles of the test specimen. To evaluate the wear property of the test specimens under rolling cycles, some wear tests were ended after 100,000; 300,000; or 500,000 cycles. After the test, the hardness of the rolling contact surface was measured with a microhardness tester, and its thin-film microstructure was observed with a transmission electron microscope (TEM).
3. Test results 3.1. Effect of pre-test hardness on weight loss Fig. 2 shows the relationship between the pre-test hardness and the weight loss of the test specimens. There is almost a linear relationship between the pre-test hardness and the weight loss. Namely, the weight loss decreases as the pre-test hardness increases. Moreover, there is a close correlation between the weight loss and the carbon content. 3.2. Effect of carbon content on weight loss
Fig. 1. Dimensions and shapes of wear test specimens and schematic illustration of rolling wear test machine.
To clarify the effect of the carbon content on the weight loss, the relationship between the carbon content and the weight loss was examined. Fig. 3 shows the relationship between the carbon content and the weight loss of the test specimens in the hardness range of 385–395 Hv. There is almost a linear correlation between the carbon content and the weight loss. Namely, the weight loss of the test specimens decreases as the carbon content increases.
M. Ueda et al. / Wear 253 (2002) 107–113
109
Fig. 2. Relationship between pre-test hardness and weight loss of test specimens.
Fig. 4. Relationship between number of rolling cycles and weight loss of test specimens with pre-test hardness range from 385 to 395 Hv.
Thus, it can be seen that the rolling contact wear property of the pearlitic steels greatly depend not only on the pre-test hardness but also on the carbon content.
It has been reported that the rolling contact wear property of rail steels is closely correlated with the number of rolling cycles and the post-test rolling contact surface hardness (RCSH) [10,11]. Accordingly, the authors investigated the relationship between the number of rolling cycles and the weight loss, and the relationship between the number of rolling cycles and the post-test RCSH by using test specimens with a pre-test hardness in the range from 385 to 395 Hv. The relationship between the number of rolling cycles
in the range from 100,000 to 700,000 and the weight loss of the test specimens is shown in Fig. 4. The relationship between the number of rolling cycles in the same range as mentioned and the post-test RCSH of the test specimens is shown in Fig. 5. There is little difference in weight loss in all of the test specimens up to 100,000 cycles. As the number of rolling cycles exceed 100,000, the weight loss changes in accordance with the carbon content, and an increase in the weight loss of the test specimens with a high carbon content is suppressed. As the number of rolling cycles increase further, the test specimens with a high carbon content have less weight loss compared with the test specimens with a low carbon content. The post-test RCSH in all the test specimens increase as the number of rolling cycles increase up to 100,000. As the number of rolling cycles exceed 100,000, the post-test
Fig. 3. Relationship between carbon content and weight loss of test specimens with pre-test hardness range from 385 to 395 Hv.
Fig. 5. Relationship between number of rolling cycles and post-test rolling contact surface hardness of test specimens with pre-test hardness range from 385 to 395 Hv.
3.3. Effects of number of rolling cycles on weight loss and post-test rolling contact surface hardness (RCSH)
110
M. Ueda et al. / Wear 253 (2002) 107–113
Fig. 6. TEM images beneath rolling contact surface of test specimens and selected area diffraction patterns after 700,000 cycles: (a) bright-field image of 0.8 mass%C steel; (b) dark-field image of 0.8 mass%C steel; (c) bright-field image of 1.0 mass%C steel; (d) dark-field image of 1.0 mass%C steel.
RCSH changes in accordance with the carbon content, and an increase in the post-test RCSH of the test specimens with a high carbon content is promoted. As the number of rolling cycles increase further, the post-test RCSH of the test specimens with a high carbon content is higher than that of the test specimens with a low carbon content. When the number of rolling cycles is 700,000, the post-test RCSH increases to 670 Hv for the 0.8 mass%C steel, to 710 Hv for the 0.9 mass%C steel, and to 750 Hv for the 1.0 mass%C steel.
the small area. Upon considering the difference between the carbon content and the size of crystalline grains in the dark field images, it is noted that the size of crystalline grains of the test specimen with a high carbon content is smaller than that of the test specimen with a low carbon content. This means that the matrix ferrite in the rolling contact surface is refined due to the increased carbon content of the pearlitic steels.
3.4. Observation of microstructure beneath rolling contact surface
4. Discussions
Fig. 6 shows TEM images beneath the rolling contact surface of the test specimens of the 0.8 and 1.0 mass%C steel after 700,000 cycles and the diffraction patterns obtained from a selected area of 1 mm in diameter. In the bright field images, the fracturing in the cementite phase, dislocations, and some dislocation cells are observed in the matrix ferrite in each test specimen. In the diffraction patterns of the photographed regions, diffraction spots of the cementite phase and the ferrite phase are dispersed, and the diffraction spots of the ferrite phase are formed in rings in each test specimen. This means that there are a lot of crystalline grains which are thought to be subgrains or nanocrystalline grains within
These test results showed that the change in the weight loss is closely correlated with the RCSH and that the RCSH depends on the number of rolling cycles in the wear test and the carbon content of the test specimens. In this chapter, the authors quantify the relationship between the RCSH and weight loss, and clarify the dominating factor in the rolling contact wear in pearlitic steels and the effects of carbon content on this factor. Further, the authors discuss the mechanism whereby the microstructure beneath the rolling contact surface varies as carbon content increases and the work-hardening rate of the rolling contact surface from the wear test results and the observation results of rolling contact surfaces.
M. Ueda et al. / Wear 253 (2002) 107–113
Fig. 7. Relationship between average rolling contact surface hardness and wear rate of test specimens at 200,000 cycle intervals between 100,000 and 700,000 cycles.
4.1. Dominating factor in rolling contact wear in pearlitic steels and effects of carbon content on this factor To clarify the relationship between the RCSH and the weight loss, the authors examined the relationship between the average RCSH and the wear rate in a range of rolling cycles where the RCSH and the weight loss clearly differed in accordance with the carbon content. Fig. 7 shows the relationship between the average RCSH and the wear rate (W) of test specimens at 200,000 cycle intervals between 100,000 and 700,000 cycles. There is almost a linear relationship between the average RCSH and the wear rate, and the carbon content is not a factor therein. Namely, the weight loss of pearlitic steel decreases as the RCSH increases. To confirm the effect of the carbon content on the RCSH, the authors, noting the difference between the pre- and post-test RCSH, set the ratio of the increase in the post-test RCSH to the pre-test hardness. The authors referred to this as the work-hardening rate of the rolling contact surface shown by Eq. (1), and examined the relationship between the carbon content and the work-hardening rate of the rolling contact surface. Fig. 8 shows the relationship between the carbon content of the test specimens and the work-hardening rate of the rolling contact surface. There is a little change in the work-hardening rate of the rolling contact surface at 100,000 cycles, at which the RCSH does not change as carbon content increases. However, there is a close relationship between the carbon content and the work-hardening rate of the rolling contact surface at 700,000 cycles, at which the RCSH rises as carbon content increases. That is, the work-hardening rate of the rolling contact surface rises as the carbon content increases after a
111
Fig. 8. Relationship between carbon content and work-hardening rate of rolling contact surface of test specimens.
period of rolling contact. Work-hardening rate of rolling contact surface (%) (post-test RCSH − pre-test hardness) × 100 = (pre-test hardness)
(1)
where the hardness values are all expressed in Vickers hardness (Hv). From these results, it is concluded that the dominating factor in the rolling contact wear of pearlitic steels is the RCSH and that the increase in the RCSH depends on the rise in the work-hardening rate of the rolling contact surface as the carbon content increases. Namely, it is thought that the improved wear resistance of pearlitic steels is attributable to an increase in RCSH due to raising the work-hardening rate of the rolling contact surface as the carbon content increases. 4.2. Mechanism for change in microstructure beneath rolling contact surface and rise in work-hardening rate of rolling contact surface with increase in carbon content According to these test results and observations, the reason why the work-hardening rate of the rolling contact surface of pearlitic steels rises is considered to be as follows: The rolling contact introduces strain into the ferrite phase and at the same time produces minute fractures in the cementite phase beneath the rolling contact surface. Repeated rolling contact concentrates strain in the ferrite phase that is lower in hardness than the fractured cementite. The strain concentration forms many dislocations in the matrix ferrite, and promotes dislocation hardening and grain refinement in the matrix ferrite. As a result, the matrix ferrite is strengthened by dislocation hardening and grain refinement. Moreover, the mechanism whereby the work-hardening rate of the rolling contact surface rises as the carbon content of the pearlitic steels increases is explained as follows: The
112
M. Ueda et al. / Wear 253 (2002) 107–113
increase in the carbon content of the pearlitic steels raises the density of the hard cementite phase. The increase in the cementite density promotes the concentration of strain in the matrix ferrite, and accelerates the grain refinement in the matrix ferrite. As a result, the matrix ferrite is strengthened further through the promotion of dislocation hardening and grain refinement as the carbon content increases. Tarui et al. reported that when a high carbon wire is cold-drawn, the cementite phase in the pearlite structure is decomposed, and the carbon of the cementite phase is dissolved into the matrix ferrite phase, and the ferrite phase is strengthened by the solid solution of carbon [12]. The rolling contact surface is subject to a more severe processing condition than that of the wire drawing, and the cementite phase is presumed to be decomposed by the rolling contact. Therefore, the increased amount of carbon dissolving into the matrix ferrite as the cementite density increases is considered to be one of the factors responsible for the raised work-hardening rate of the rolling contact surface.
5. Status of trial rails installation Based on the results of the laboratory test described, it was confirmed that the wear resistance in pearlitic steels could be improved by increasing the carbon content of these steels. To confirm the properties of rails in actual use, therefore, pearlitic steel rails with a carbon content of 0.9 mass% and hardness of HB 370 were produced on a trial basis and installed in actual tracks. In this chapter, one example of the wear property of these pearlitic steel rails on heavy-haul tracks in North America is described. The new rails were tested in an actual track and compared with the conventional pearlitic steel rails with a carbon content of 0.8 mass% and hardness of HB 370 (called DHH370). Fig. 9 shows the relationship between the passing tonnage
Fig. 9. Relationship between passing tonnage and amount of head side wear of rails in a curved track.
and the amount of head side wear of these rails in a curved track having a 290 m radius. As for the new rail, the amount of head side wear was less than that of the DHH370 rail in every passing tonnage. It was confirmed that the wear resistance was improved by increasing the carbon content of the pearlitic steels in the actual track. Moreover, when the service life of the new rail was compared with the DHH370 rail in half inches in the head side wear, it was clear that the service life of the new rail had been improved by about 55%. Upon confirming the improvement of the service life in actual tracks, the authors developed the Hyper-Eutectoid steel rails (named HE370, HE400) with the carbon content of 0.9 mass% and hardness of HB 370 and HB 400. It is thought that these rails will contribute in the future to greatly improving the service life of rails in heavy-haul railways where excellent wear resistance is demanded.
6. Conclusions Pearlitic steels with different carbon contents were two-cylinder wear tested to study the effect of the carbon content on the rolling contact wear. The dominating factor in the rolling contact wear in pearlitic steels and the effects of carbon content on this factor were examined. Further, the mechanism for the change in the microstructure beneath the rolling contact surface and the rise in the work-hardening rate of the rolling contact surface as the carbon content increases were discussed. The main findings obtained are as follows: (1) The wear resistance of pearlitic steels improve as the carbon content increases. (2) The dominating factor in the rolling contact wear of pearlitic steels is the RCSH. That is, the wear resistance of pearlitic steels improves as the RCSH increases. (3) The increase in the RCSH depends on the rise in the work-hardening rate of the rolling contact surface as the carbon content increases. (4) The improving wear resistance of pearlitic steels is attributable to an increase in the RCSH due to raising the work-hardening rate of the rolling contact surface as the carbon content increases. (5) The reason why the work-hardening rate of the rolling contact surface of pearlitic steel rises as the carbon content increases is considered to be as follows: an increase in the cementite density increases the amount of dislocation in the matrix ferrite and promotes the grain refinement of the matrix ferrite. As a result, the matrix ferrite is strengthened through the promotion of dislocation hardening and grain refinement. (6) Newly developed Hyper-Eutectoid steel rails (HE370) with a carbon content of 0.9 mass% show an improvement in service life of about 55% as compared with the conventional pearlitic steel rail (DHH370) with a carbon content of 0.8 mass% in an actual track.
M. Ueda et al. / Wear 253 (2002) 107–113
Acknowledgements The authors wish to thank Union Pacific Railroad and Burlington Northern Santa Fe for permitting the actual track tests to be conducted and Dr. Senuma for the fruitful discussions with him and his helpful comments. References [1] G.A. Tomlinson, Philos. Mag. 7 (1927) 905–911. [2] M. Ishida, Shinsenro 49 (1995) 38–40. [3] H. Masumoto, K. Sugino, H. Hayashida, in: Proceedings Conference on First International Heavy Haul Railway, Session 212, Colorado Springs, 1978. [4] K. Sugino, H. Kageyama, H. Masumoto, in: Proceedings Conference on Second International Heavy Haul Railway, Perth, 1982, pp. 187–197.
of the IHHA, of the IHHA,
113
[5] K. Sugino, H. Kageyama, C. Urashima, in: Proceedings of the 32nd Mechanical Working and Steel, ISS, Warrendale, 1990, 171–176. [6] P. Clayton, N. Jin, Wear 200 (1996) 74–82. [7] H. Yokoyama, S. Yamamoto, M. Fujikake, Y. Yoshida, in: Proceedings of the Fourth International Symposium on Rail Steels, ISS, Warrendale, 1997, pp. 1023–1028. [8] M. Ueda, K. Uchino, H. Kageyama, K. Kutaragi, K. Babazono, in: Proceedings of the Conference on Sixth International Heavy Haul Railway, IHHA, Cape Town, 1997, pp. 355–369. [9] M. Ueda, K. Uchino, H. Kageyama, M. Kaneta, A. Kobayashi, in: Proceedings of the Conference on IHHA’99 STS, Vol. 1, IHHA, Moscow, 1999, pp. 259–266. [10] K. Sugino, H. Masumoto, S. Nishida, C. Urashima, H. Kageyama, M. Hattori, Seitetsukenkyu 303 (1980) 23–38. [11] W.R. Tyfour, J.H. Beynon, A. Kupoor, Wear 180 (1995) 79– 89. [12] K. Tarui, N. Maruyama, G. Shigesato, H. Tashiro, Zairyou-to-Process 11 (1998) 1327.