- Email: [email protected]
The influence of microstructure on the rolling contact fatigue of steel for high-speed-train wheel
Wear 342-343 (2015) 349–355
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
The influence of microstructure on the rolling contact fatigue of steel for high-speed-train wheel Gen Li, Zhiyuan Hong, Qingzhi Yan n Institute of Advanced Ceramics and Nuclear Materials, Department of Material Science and Engineering, University of Science and Technology, Beijing, 30 Xueyuan Road, Haidian District, Beijing 100083, People's Republic of China
art ic l e i nf o
a b s t r a c t
Article history: Received 9 June 2015 Received in revised form 16 September 2015 Accepted 1 October 2015 Available online 14 October 2015
Rolling contact fatigue (RCF) is an important concern for the durability of railroad wheels and rails, especially when considering high-speed trains. In this work, the effect of microstructures on the RCF of high-speed-railway was studied under a simulated train speed of 500 km/h. Two different carbon steel microstructures: spheroidized pearlite and lamellar pearlite, were achieved by different heat treatment. Results of rolling contact fatigue test demonstrated that the fatigue resistance of lamellar steel considerably exceeds that of spheroidized steel. The lamellar steel presented a more fragmentary surface but lower weight loss during the entire fatigue tests. Analysis of cross-section profiles of worn specimens suggests that cracks propagated near the surface of lamellar steel and may strengthen the fatigue resistance of the steel. Based on micro-indentation hardness data, it is suggested that greater energy dissipation induced by larger plastic deformation may also improve RCF resistance. & 2015 Elsevier B.V. All rights reserved.
Keywords: Rolling contact fatigue Steel Rail–wheel tribology Surface topograghy
1. Introduction Taken a variety of aspects such as capacity, cost, speed and safety into account, railway, especially high-speed-railway, is an outstanding means of transportation. A considerable amount of money and time have been devoted into projects focusing on improving the speed and safety of high-speed-railway all over the world. The service life of wheel and rail plays a critical role in operation of high-speed-railway and naturally catches researchers' eyes in different disciplines. Wear and rolling contact fatigue (RCF) are considered to be two of the main failure mechanisms for wheel and rail. A lot of investigations have been made in those two fields. Many researches have been done in diverse aspects of wheel materials. Jha [1] focused on the effect of microstructure on the abrasion resistance of steels and proposed that combining of soft ferrite and hard martensite leads to a better abrasion resistance. Deters [2] investigated the wear behavior of the wheel and rail and discussed the influence of load, creepage and speed, found that the wear volume decreased as the running speed increased but presented opposite behaviors as the load and creepage rose. Garnham [3,4] focused on the RCF of rail and wheel and studied the influence of different operating conditions (like creepage and load) and microstructures of materials on RCF of the rails. Ekberg [5] fully described the n
Corresponding author E-mail address: [email protected] (Q. Yan).
http://dx.doi.org/10.1016/j.wear.2015.10.002 0043-1648/& 2015 Elsevier B.V. All rights reserved.
fatigue of wheel and rail and discussed the influence factor of the fatigue. Furthermore, Benuzzi [6] noticed the competitive relationship between wear and RCF and investigated the propagation of the surface cracks. In 1980s, RCF had drawn the attention of European Rail Research Institute; thus experiments of distinct scale and numerical simulations had been used to explore this problem [7]. Nelias [8] studied the RCF performance of bearing steels with artificial dents under different working condition on a high-speed twin-disk machine and found that sliding creepage had a great influence on the durability of the materials. Fletcher [9] conducted full-scale experiment and numerical simulation to investigate the fluid penetration of cracks under walk pace and suggested that the growth rate of fatigue crack was around 10 times the rail wear rate. Ringsberg [10] proposed a strategy combining elastic–plastic finite element (FE) analyses, multiaxial fatigue crack initiation models used together with the critical plane concept, fatigue damage summation calculations to predict the fatigue life of rolling contact fatigue crack initiation. In those previous researches, pearlitic steel is commonly used for manufacturing wheel and rail due to its great wear resistance and appropriate ductility [11–13]. Among the pearlitic steels with different microstructures (metallographic structure), spheroidized pearlitic steel is the most popular material for wheel due to the low cost and easy fabrication. Some researchers [14] found that the wear resistance of lamellar pearlitic steel is better than spheroidized pearlitic steel because of the more effective work hardening of cementite lamellar. However, the influence of different microstructure on the rolling contact fatigue, and eventually
350
G. Li et al. / Wear 342-343 (2015) 349–355
the service life, of wheel under high speed is still unknown even though the problem of rolling contact fatigue is fully developed. This work intends to achieve two kinds of pearlitic steels with same hardness and components but different microstructures and explore the influence of microstructures on the rolling contact fatigue performance. Furthermore, most of existing researches were performed at a relatively lower speed. Since the high-speed-rail in China has already been running above 300 km/h, the operation of rail/wheel under even higher speed should be appropriate for the purpose of safety. Thus the study of wear and fatigue behavior of materials under a high speed is needed, which is not mentioned in the previous paper. Besides, though the wear of the wheel and rail decreased with the increase of running speed [2], the RCF of railway increased a lot with the rise of speed. Thus, serving life of wheel at extremely high speed is unclear. In the present research, a series of twin-disc experiments were conducted to simulate the speed of the wheel at 500 km/h. Through this research, the author expects to find the fatigue behavior of materials under extremely high speed and the influence of microstructure on the service life of wheel.
2. Material and methods 2.1. Materials The components of the steel are listed in Table 1. The steel was fabricated through vacuum induction melting. Then the steel ingot was held at 1523 K for 90 min and rolled to a plate of 15 mm between 1223 and 1398 K. Next, two groups of specimens with a diameter of 59.5 mm were cut from the plate. Specimen A was held at 1113 K for 20 min and water quenched to 813 K and then insulated for 120 min and finally cooled down in air. Specimen B was held at 1113 K for 20 min and water quenched to 923 K and spray cooled to room temperature. Another kind of pearlitic steel was used to simulate the rail in this experiment. The components of the rail are also listed in Table 1.
2.4. Microstructure examinations The profiles of original and worn specimen are first ground and polished through the same procedure and then eroded by the alcoholic solution containing 3% of nitric acid (volume fraction). Then, the microstructure of the specimen is examined with a scanning electron microscope (SEM). The worn surface of the wheel was also examined under SEM (FEI Quanta 450, America). 2.5. Microhardness tests The microhardness test was conducted in accordance with GB/T 4340.1-2009 on microhardness tester. The cross-section profiles of specimens were ground and polished through the same procedure either. The data of different distances below the surface were repetitively tested 3 times and an average value was calculated.
3. Results 3.1. Microstructure and mechanical properties of steels
2.2. Tensile tests Tensile experiments were conducted in accordance with GB/T 228-2002 on MTS809 tensile test machine (MTS, America) with the tensile rate of 2 mm/min. The tensile specimens were cut into Ф5 25 mm2. 2.3. Rolling contact fatigue experiments The fatigue experiment of wheel and rail was undertaken in accordance with YB/T 5345-2006 on a twin-disc MJP-20 RCF test machine (China). The as-treated sample was located at the top simulating the wheel, while the sample of a diameter of 60.5 mm was located at the bottom simulating the rail in the test. The geometry of the specimen is listed in Fig. 1. In order to accelerate the propagation of the cracks [5] and maintain a suitable temperature during the test, HL32 hydraulic liquid was introduced in this test. The kinematic coefficient of viscosity, flash point and density of this Table 1 Component of wheel and rail steel.
Wheel Rail
liquid were 32 mm2/s, 493 K and 0.875 g/cm3 respectively. In order to ensure the reliability of the experiment results, rolling contact fatigue tests under three distinct conditions were conducted. The creepage, rotation speed and normal pressure of these tests are set as follows: 1) 1.6%, 2200 rpm, 1000 MPa; 2) 1.6%, 2200 rpm, 1200 MPa; 3) 1.6%, 2800 rpm, 1000 MPa. The frequency of 2200 rpm during the test is the same with that of the realistic running speed of 400 km/h, while the frequency of 2800 rpm equals that of the realistic running speed of 500 km/h. Meanwhile, the normal pressure of 1000 MPa is the same with the realistic pressure under the wheel/ rail contact and the performance of the materials under higher pressure (1200 MPa) was also investigated. The specimen was cleaned, dried, weighed and carefully examined after each 500 thousands of rotations. According to YB/T 5345-2006, a vibration transducer was assembled to detect the vibration over 0.03 mm, which is induced by the fatigue pit of the steel surface. Once the vibration occurred, the test machine would stop immediately.
C
Si
Mn
P
S
Cr
Fe
0.4 0.7
0.35 0.26
0.75 1.13
o 0.02 o 0.02
o 0.02 o 0.02
0.25 \
Bal. Bal.
The images of original and heat-treated steels (spheroidized and lamellar) are presented in Fig. 2. It can be seen from the Fig. 2 (b) that tiny spheroidized carbide in specimen A, the tempered pearlitic steel, uniformly distributed on ferrite matrix. The microstructure of pearlite in specimen B, the water sprayed pearlitic steel, presented a totally different morphology. Typical lamellar pearlite with lamellar spacing of around 0.3 μm and little pre-eutectoid ferrite are demonstrated in Fig. 2(c). Through the comparison of the original microstructure in Fig. 2(a) with the following two microstructures, it can be concluded that both two heat treatments have avoided the eutectoid of large ferrite at the boundary of austenite. The mechanical properties of original and heat treated steels are listed in Table 2. The tensile strength, both ultimate tensile strength and yield tensile strength, of two different steels had risen greatly after the heat treatment. The ultimate tensile strength of both two steels was elevated to above 900 MPa, while the yield tensile strength of spheroidized steel and lamellar steel had risen by 326.4 and 121.9 MPa respectively. Meanwhile, the elongation of steel, the symbol of ductility, had not decreased much. Those improvements of mechanical properties of steels attribute to the little pre-eutectoid ferrite and thin lamellar spacing, which avoid the weak zone at the boundary of austenite and improve the ductility. Besides, the surface hardness of both two
G. Li et al. / Wear 342-343 (2015) 349–355
351
Fig. 1. The geometry of specimen: (a) wheel; (b) rail.
Fig. 2. Microstructure of original steel and specimen A and B: (a) original steel; (b) specimen A, tempered pearlitic steel; (c) specimen B, water sprayed pearlitic steel.
Table 2 Mechanical properties of three steels. Steels
Original Quenched-tempered (spheroidized steel) Water-sprayed cooling (lamellar steel)
Table 3 Experiment parameters of rolling contact fatigue.
Ultimate tensile strength
Yield tensile Elongation Hardness strength
(MPa)
(MPa)
(%)
(HRC)
849.7 955.6
545.2 871.6
18.9 15.2
26.7 32
903.5
667.1
17.6
32
steels was controlled to 32 HRC, which eliminated the influence of surface hardness to fatigue behavior. 3.2. Rolling contact fatigue test Due to the great improvements of mechanical properties after the heat treatment, the author of this work focused on the fatigue behaviors of heat-treated steels and the fatigue tests were only conducted on those two heat-treated steels. According to the YB/T 5345-2006, once the area of deep flaking exceeds 3 mm2 or more, the materials are considered to be failure. However, if the wheels maintain its properties after 1*107 stress cycles, the materials are considered to be fatigue-free and experiments can be ended. The parameters and results of rolling contact fatigue tests are displayed in Tables 3 and 4. It can be concluded that both two kinds of pearlitic steels presented excellent rolling contact fatigue resistance under relatively low speed and pressure. However, lamellar pearlitic steels demonstrated better fatigue resistance under higher speed (2800 rpm) and higher pressure (1200 MPa). Since the surface and cross-section morphology of both two kinds
Group
Creepage (%)
Rotation speed (rpm)
Normal pressure (MPa)
1 2 3
1.6 1.6 1.6
2200 2200 2800
1000 1200 1000
of steels is similar, the following description and discussion are going to focus on the specimen in group 3. During the fatigue test, the quenched-tempered pearlitic steel (specimen A) presented two deep pits after 4.23*106 stress cycles and the areas of two pits were 2 mm2, while the water-sprayed cooling pearlitic steel remained almost complete after 107 stress cycles (Fig. 3). Fig. 4 demonstrates the weight loss after each 5*105 stress cycles. It can be concluded that the weight loss of quenchedtempered steel was always larger than water-sprayed steel. Besides, the weight loss rate of water-sprayed steel remained nearly unchanged, while the weight loss rate of quenchedtempered steel kept rising during the test. 3.3. Worn surface of the wheel Fig. 5 demonstrates the worn surface of the specimens after fatigue experiments. The worn surface of quenched-tempered steel was more complete but the holes on the surface were much deeper than the water-sprayed steel. The worn surface of water-sprayed cooling steel, where presented a number of shallow flat hollows, suggested that lots of flake debris had been removed from the surface. Besides those holes, there existed a lot of scratches on the surfaces of both two steels, which was certainly
352
G. Li et al. / Wear 342-343 (2015) 349–355
induced by hard particles on counterfeit rails or the third body between the rail and wheel. 3.4. Profile of the specimen after rolling contact fatigue test Fig. 6 displays the typical profiles of the specimens after the rolling contact fatigue test. It can be seen from Fig. 6(a) that there lay a number of dark dots below the surface of quenchedtempered steel. Most of these dark dots were located within 20 μm deep. Since the dark regions under electronic microscope represent the lower region of the profile after erosion, these dark Table 4 Results of rolling contact fatigue tests. Group Material
Lifespan 105 cycles
Description
1
1000
No fatigue was observed
2
3
Quenched-tempered steel Water-sprayed steel Quenched-tempered steel Water-sprayed steel Quenched-tempered steel Water-sprayed steel
1000 568 1000 423 1000
Fatigue pits presented No fatigue was observed Fatigue pits presented No fatigue was observed
dots might be the dislocation pile-up groups. Because of the stress introduced during the fatigue test, dislocations began to glide or climb and finally piled up at the obstacles. Those dislocation pileup groups would be easily eroded and leave a dark region at the SEM image. However, little such dark dots existed at a deeper area. Profile of water-sprayed steed presented by Fig. 6(b) displays some different features. It is clear that the lamellar pearlite within 5 μm under the surface had been turned to be parallel with the surface. This direction is also the sliding direction. In this rotating area, some cracks and flaws were observed. Besides, the closer to the surface, the smaller the distance of the lamellar spacing of the pearlite would be. This phenomenon is much clearer in Fig. 7, which demonstrates the comparison of lamellar spacing at 30 μm below the surface of the tested specimen and original specimen. It can be concluded that lamellar spacing decreased from 0.3 μm to 0.15 μm at the 10 μm below the contact surface. 3.5. Micro-hardness results The micro hardness results are demonstrated in Fig. 8. Since the original hardness of both two steels is 32 HRC, which is corresponding to 319 HV in Vicker's hardness, it can be concluded that the working hardening zone of quenched-tempered pearlitic steel is around 150 μm below the surface, while that of water-sprayed cooling pearlitic steel is 300 μm approximately. From Fig. 7, it can also be drawn that the magnitude of working hardening of lamellar pearlitic steel is larger than spheroidized pearlitic steel. This
Fig. 3. Macromorphology of two pearlitic steels after fatigue test: (a) quenched-tempered steel and (b) water-sprayed steel.
G. Li et al. / Wear 342-343 (2015) 349–355
conclusion is also proved by the surface hardness of two steels. The surface hardness of worn spheroidized pearlitic and lamellar pearlitic steel is 356.4 and 395.3 HV respectively.
4. Discussion Earlier paper [5] concluded that most of the cracks initiated at the surface of the wheel or at the sub-surface of the wheel, where the maximum shear stress occurred. It also suggested that the maximum shear stress often occurred at around 4–5 mm below the surface. In this research, no cracks were observed at that area of both two steels. In fact, no cracks were detected at
Fig. 4. Weight losses of two specimens.
353
1 mm or more below the surface. Therefore, fatigue behavior and service life of wheel were mainly determined by the surface initiated cracks. From the experimental data of rolling contact fatigue test, lamellar pearlitic steel demonstrated a better fatigue resistance than spheroidized pearlitic steel. Such behavior may attributed to the different interaction between wear and fatigue for those two different steels. The interaction between wear and fatigue, also known as tribo-fatigue [15,16], proposed that the damage caused by wear and fatigue cannot be simply added when the materials were undertaking both wear and fatigue. Instead, the wear may reduce the fatigue damage. Some researchers [11,12,14,17] have proposed that lamellar pearlitic steel with small lamellar spacing demonstrates great wear resistance because thin cement sheet show better ductility than spheroidized cement. In this research, the plastic deformation area of lamellar pearlitic steel is much larger than the spheroidized pearlitic steel. Thin lamellar perlite with better ductility, together with larger plastic deformation area, would certainly relieve more stress concentration and dissipate more energy. Naturally, with more energy dissipated, the propagation speed of cracks in lamellar pearlitic steel would be slower. On the other hand, the different worn surface morphology and distinguishing mechanical properties of two different pearlitic steels suggest totally different tribo-fatigue behaviors. Although the thin lamellar structure improves the ductility of pearlite, such structure also increases the magnitude of working hardening of ferrite lamella [14,17]. The thin ferrite lamella experienced larger strain than the ferrite matrix in spheroidized pearlite and presented larger working hardening (Fig. 8). Thus, ductile exhaustion of thin ferrite lamella might be induced by such severe plastic deformation and the ferrite region became the weak region
Fig. 5. Worn surface of two specimens: (a) quenched-tempered steel; (b) water-sprayed steel.
Fig. 6. Typical profiles of the specimens after fatigue test: (a)quenched-tempered steel; (b) water-sprayed steel.
354
G. Li et al. / Wear 342-343 (2015) 349–355
Fig. 7. Lamellar spacing of original water-sprayed steel (a) and steel after fatigue test (b).
Fig. 8. Micro-hardness of two steels against distance below the surface.
eventually attenuates the fatigue damage and reduces the weight loss of materials. Instead of this above-mentioned crack propagation process in lamellar steel, the propagation of cracks in spheroidized steel presented a typical propagation process of fatigue cracks. The cracks grew into the material at a shallow angle and then deviated to the sliding direction of the wheel. Then, the crack deviated again to the surface of material and left a deep hole on the surface of the wheel. Fig. 9 demonstrates a crack that almost grew to the surface again. Such a propagation process was occurred at much deeper area below the surface; thus the wear process had little influence on this fatigue process.
5. Conclusions Fig. 9. Crack propagation of spheroidized steel.
of materials. When the surface initiated crack propagated to this weak region, the crack may deviated to this region and propagated along this lamellar region. Once the crack intersected with another surface initiated crack, a flake of debris was removed from the surface. Since the near-surface area (0–5 μm below the surface) had suffered a large amount of plastic deformation, as presented in Fig. 5 (those defects in this region), the above-mentioned propagation of cracks might occurred in this region with a high possibility. The flat shallow pits on the surface give evidence of this crack propagation process. However, this propagation process occurred at the place that extremely close to the surface, where wear damage may also happened. Once the region was removed by wear damage, such fatigue damage was “eliminated” by wear damage. Such competitive rather than superimposed relationship
Two kinds of pearlitic steels with different microstructures were achieved by distinct heat treatment in this work. Tensile strength of both two steels increased a lot while the elongation of materials decreased a little after the heat treatment. The spheroidized steel presented a higher strength but the surface hardness for both two steels was maintained at the same value of 32 HRC. Rolling contact fatigue tests of both two steels suggested that the spheroidized steel failed at 4.23*106 cycles with two deep fatigue pits of around 2 mm2, while the lamellar steel allowed no obvious pits after 107 cycles. Meanwhile, weight loss of spheroidized steel is larger than lamellar steel during the fatigue tests. Those results demonstrated that lamellar steel has better fatigue resistance. The worn surface examination of both two steels displayed that the spheroidized steel maintained a more complete surface with some deep pits while the lamellar one presented a more fragmentary morphology but only with some shallow flake pits. Morphology of
G. Li et al. / Wear 342-343 (2015) 349–355
profile for lamellar steel indicates that the surface initiated cracks propagate along the direction of rolling in the near-surface region (less than 5 μm). Such a special propagation process was induced by the plastic exhaustion of lamellar ferrite and would lead to an unusual tribo-fatigue behavior. The wear on the surface of materials may eliminate the fatigue cracks in the near-surface and thus elevate the fatigue resistance of lamellar steel. Moreover, microhardness tests of the experimented specimens suggest that lamellar steel holds a larger plastic deformation zone. Thus more energy would be delivered to plastic deformation and naturally, less energy would be delivered to cracks propagation. Combining those two aspects, lamellar steel presents better fatigue resistance. Tribo-fatigue view in this paper provides a novel supplement for RCF problem and the results of this research suggest a promising future for lamellar steel equipped as wheel material for high-speed-railway.
Acknowledgments This work was supported by a grant from the Major State Basic Research Development Program of China (973 Program) (No. 2011CB711103).
References [1] A.K. Jha, B.K. Prasad, O.P. Modi, S. Das, A.H. Yegneswaran, Correlating microstructural features and mechanical properties with abrasion resistance of a high strength low alloy steel, Wear 254 (1–2) (2003) 120–128. [2] L. Deters, M. Proksch, Friction and wear testing of rail and wheel material, Wear 258 (7) (2005) 981–991. [3] J.H. Beynon, J.E. Garnham, K.J. Sawley, Rolling contact fatigue of three pearlitic rail steels, Wear 192 (1–2) (1996) 94–111. [4] J.E. Garnham, C.L. Davis, Very early stage rolling contact fatigue crack growth in pearlitic rail steels, Wear 271 (1–2) (2011) 100–112.
355
[5] A. Ekberg, E. Kabo, Fatigue of railway wheels and rails under rolling contact and thermal loading—an overview, Wear 258 (7–8) (2005) 1288–1300. [6] D. Benuzzi, E. Bormetti, G. Donzella, Stress intensity factor range and propagation mode of surface cracks under rolling–sliding contact, Theor. Appl. Fract. Mech. 40 (1) (2003) 55–74. [7] D.F. Cannon, H. Pradier, Rail rolling contact fatigue Research by the European Rail Research Institute, Wear 191 (1–2) (1996) 1–13. [8] D. Nelias, C. Jacq, G. Lormand, G. Dudragne, A. Vincent, New methodology to evaluate the rolling contact fatigue performance of bearing steels with surface dents: application to 32CrMoV13 (nitrided) and M50 steels, J. Tribol. 127 (3) (2005) 611–622. [9] D.I. Fletcher, P. Hyde, A. Kapoor, Modelling and full-scale trials to investigate fluid pressurisation of rolling contact fatigue cracks, Wear 265 (9–10) (2008) 1317–1324. [10] J.W. Ringsberg, Life prediction of rolling contact fatigue crack initiation, Int. J. Fatigue 23 (7) (2001) 575–586. [11] P. Clayton, The relations between wear behaviour and basic material properties for pearlitic steels, Wear 60 (1) (1980) 75–93. [12] J. Kalousek, D.M. Fegredo, E.E. Laufer, The wear resistance and worn metallography of pearlite, bainite and tempered martensite rail steel microstructures of high hardness, Wear 105 (3) (1985) 199–222. [13] R. Kuziak, T. Zygmunt, A new method of rail head hardening of standard-gauge rails for improved wear and damage resistance, Steel Res. Int. 84 (1) (2013) 13–19. [14] Y. Wang, T. Lei, J. Liu, Tribo-metallographic behavior of high carbon steels in dry sliding: II. Microstructure and wear, Wear 231 (1) (1999) 12–19. [15] L.A. Sosnovskiy, Tribo-Fatigue: Wear-Fatigue Damage and its Prediction, Springer Science & Business Media, Berlin, Heidelberg, 2010. [16] L.A. Sosnovskiy, Surprises of tribo-fatigue, in: Proceedings of the World Tribology Congress III, American Society of Mechanical Engineers, 2005. [17] J.E. Garnham, C.L. Davis, The role of deformed rail microstructure on rolling contact fatigue initiation, Wear 265 (9–10) (2008) 1363–1372.
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
The influence of microstructure on the rolling contact fatigue of steel for high-speed-train wheel Gen Li, Zhiyuan Hong, Qingzhi Yan n Institute of Advanced Ceramics and Nuclear Materials, Department of Material Science and Engineering, University of Science and Technology, Beijing, 30 Xueyuan Road, Haidian District, Beijing 100083, People's Republic of China
art ic l e i nf o
a b s t r a c t
Article history: Received 9 June 2015 Received in revised form 16 September 2015 Accepted 1 October 2015 Available online 14 October 2015
Rolling contact fatigue (RCF) is an important concern for the durability of railroad wheels and rails, especially when considering high-speed trains. In this work, the effect of microstructures on the RCF of high-speed-railway was studied under a simulated train speed of 500 km/h. Two different carbon steel microstructures: spheroidized pearlite and lamellar pearlite, were achieved by different heat treatment. Results of rolling contact fatigue test demonstrated that the fatigue resistance of lamellar steel considerably exceeds that of spheroidized steel. The lamellar steel presented a more fragmentary surface but lower weight loss during the entire fatigue tests. Analysis of cross-section profiles of worn specimens suggests that cracks propagated near the surface of lamellar steel and may strengthen the fatigue resistance of the steel. Based on micro-indentation hardness data, it is suggested that greater energy dissipation induced by larger plastic deformation may also improve RCF resistance. & 2015 Elsevier B.V. All rights reserved.
Keywords: Rolling contact fatigue Steel Rail–wheel tribology Surface topograghy
1. Introduction Taken a variety of aspects such as capacity, cost, speed and safety into account, railway, especially high-speed-railway, is an outstanding means of transportation. A considerable amount of money and time have been devoted into projects focusing on improving the speed and safety of high-speed-railway all over the world. The service life of wheel and rail plays a critical role in operation of high-speed-railway and naturally catches researchers' eyes in different disciplines. Wear and rolling contact fatigue (RCF) are considered to be two of the main failure mechanisms for wheel and rail. A lot of investigations have been made in those two fields. Many researches have been done in diverse aspects of wheel materials. Jha [1] focused on the effect of microstructure on the abrasion resistance of steels and proposed that combining of soft ferrite and hard martensite leads to a better abrasion resistance. Deters [2] investigated the wear behavior of the wheel and rail and discussed the influence of load, creepage and speed, found that the wear volume decreased as the running speed increased but presented opposite behaviors as the load and creepage rose. Garnham [3,4] focused on the RCF of rail and wheel and studied the influence of different operating conditions (like creepage and load) and microstructures of materials on RCF of the rails. Ekberg [5] fully described the n
Corresponding author E-mail address: [email protected] (Q. Yan).
http://dx.doi.org/10.1016/j.wear.2015.10.002 0043-1648/& 2015 Elsevier B.V. All rights reserved.
fatigue of wheel and rail and discussed the influence factor of the fatigue. Furthermore, Benuzzi [6] noticed the competitive relationship between wear and RCF and investigated the propagation of the surface cracks. In 1980s, RCF had drawn the attention of European Rail Research Institute; thus experiments of distinct scale and numerical simulations had been used to explore this problem [7]. Nelias [8] studied the RCF performance of bearing steels with artificial dents under different working condition on a high-speed twin-disk machine and found that sliding creepage had a great influence on the durability of the materials. Fletcher [9] conducted full-scale experiment and numerical simulation to investigate the fluid penetration of cracks under walk pace and suggested that the growth rate of fatigue crack was around 10 times the rail wear rate. Ringsberg [10] proposed a strategy combining elastic–plastic finite element (FE) analyses, multiaxial fatigue crack initiation models used together with the critical plane concept, fatigue damage summation calculations to predict the fatigue life of rolling contact fatigue crack initiation. In those previous researches, pearlitic steel is commonly used for manufacturing wheel and rail due to its great wear resistance and appropriate ductility [11–13]. Among the pearlitic steels with different microstructures (metallographic structure), spheroidized pearlitic steel is the most popular material for wheel due to the low cost and easy fabrication. Some researchers [14] found that the wear resistance of lamellar pearlitic steel is better than spheroidized pearlitic steel because of the more effective work hardening of cementite lamellar. However, the influence of different microstructure on the rolling contact fatigue, and eventually
350
G. Li et al. / Wear 342-343 (2015) 349–355
the service life, of wheel under high speed is still unknown even though the problem of rolling contact fatigue is fully developed. This work intends to achieve two kinds of pearlitic steels with same hardness and components but different microstructures and explore the influence of microstructures on the rolling contact fatigue performance. Furthermore, most of existing researches were performed at a relatively lower speed. Since the high-speed-rail in China has already been running above 300 km/h, the operation of rail/wheel under even higher speed should be appropriate for the purpose of safety. Thus the study of wear and fatigue behavior of materials under a high speed is needed, which is not mentioned in the previous paper. Besides, though the wear of the wheel and rail decreased with the increase of running speed [2], the RCF of railway increased a lot with the rise of speed. Thus, serving life of wheel at extremely high speed is unclear. In the present research, a series of twin-disc experiments were conducted to simulate the speed of the wheel at 500 km/h. Through this research, the author expects to find the fatigue behavior of materials under extremely high speed and the influence of microstructure on the service life of wheel.
2. Material and methods 2.1. Materials The components of the steel are listed in Table 1. The steel was fabricated through vacuum induction melting. Then the steel ingot was held at 1523 K for 90 min and rolled to a plate of 15 mm between 1223 and 1398 K. Next, two groups of specimens with a diameter of 59.5 mm were cut from the plate. Specimen A was held at 1113 K for 20 min and water quenched to 813 K and then insulated for 120 min and finally cooled down in air. Specimen B was held at 1113 K for 20 min and water quenched to 923 K and spray cooled to room temperature. Another kind of pearlitic steel was used to simulate the rail in this experiment. The components of the rail are also listed in Table 1.
2.4. Microstructure examinations The profiles of original and worn specimen are first ground and polished through the same procedure and then eroded by the alcoholic solution containing 3% of nitric acid (volume fraction). Then, the microstructure of the specimen is examined with a scanning electron microscope (SEM). The worn surface of the wheel was also examined under SEM (FEI Quanta 450, America). 2.5. Microhardness tests The microhardness test was conducted in accordance with GB/T 4340.1-2009 on microhardness tester. The cross-section profiles of specimens were ground and polished through the same procedure either. The data of different distances below the surface were repetitively tested 3 times and an average value was calculated.
3. Results 3.1. Microstructure and mechanical properties of steels
2.2. Tensile tests Tensile experiments were conducted in accordance with GB/T 228-2002 on MTS809 tensile test machine (MTS, America) with the tensile rate of 2 mm/min. The tensile specimens were cut into Ф5 25 mm2. 2.3. Rolling contact fatigue experiments The fatigue experiment of wheel and rail was undertaken in accordance with YB/T 5345-2006 on a twin-disc MJP-20 RCF test machine (China). The as-treated sample was located at the top simulating the wheel, while the sample of a diameter of 60.5 mm was located at the bottom simulating the rail in the test. The geometry of the specimen is listed in Fig. 1. In order to accelerate the propagation of the cracks [5] and maintain a suitable temperature during the test, HL32 hydraulic liquid was introduced in this test. The kinematic coefficient of viscosity, flash point and density of this Table 1 Component of wheel and rail steel.
Wheel Rail
liquid were 32 mm2/s, 493 K and 0.875 g/cm3 respectively. In order to ensure the reliability of the experiment results, rolling contact fatigue tests under three distinct conditions were conducted. The creepage, rotation speed and normal pressure of these tests are set as follows: 1) 1.6%, 2200 rpm, 1000 MPa; 2) 1.6%, 2200 rpm, 1200 MPa; 3) 1.6%, 2800 rpm, 1000 MPa. The frequency of 2200 rpm during the test is the same with that of the realistic running speed of 400 km/h, while the frequency of 2800 rpm equals that of the realistic running speed of 500 km/h. Meanwhile, the normal pressure of 1000 MPa is the same with the realistic pressure under the wheel/ rail contact and the performance of the materials under higher pressure (1200 MPa) was also investigated. The specimen was cleaned, dried, weighed and carefully examined after each 500 thousands of rotations. According to YB/T 5345-2006, a vibration transducer was assembled to detect the vibration over 0.03 mm, which is induced by the fatigue pit of the steel surface. Once the vibration occurred, the test machine would stop immediately.
C
Si
Mn
P
S
Cr
Fe
0.4 0.7
0.35 0.26
0.75 1.13
o 0.02 o 0.02
o 0.02 o 0.02
0.25 \
Bal. Bal.
The images of original and heat-treated steels (spheroidized and lamellar) are presented in Fig. 2. It can be seen from the Fig. 2 (b) that tiny spheroidized carbide in specimen A, the tempered pearlitic steel, uniformly distributed on ferrite matrix. The microstructure of pearlite in specimen B, the water sprayed pearlitic steel, presented a totally different morphology. Typical lamellar pearlite with lamellar spacing of around 0.3 μm and little pre-eutectoid ferrite are demonstrated in Fig. 2(c). Through the comparison of the original microstructure in Fig. 2(a) with the following two microstructures, it can be concluded that both two heat treatments have avoided the eutectoid of large ferrite at the boundary of austenite. The mechanical properties of original and heat treated steels are listed in Table 2. The tensile strength, both ultimate tensile strength and yield tensile strength, of two different steels had risen greatly after the heat treatment. The ultimate tensile strength of both two steels was elevated to above 900 MPa, while the yield tensile strength of spheroidized steel and lamellar steel had risen by 326.4 and 121.9 MPa respectively. Meanwhile, the elongation of steel, the symbol of ductility, had not decreased much. Those improvements of mechanical properties of steels attribute to the little pre-eutectoid ferrite and thin lamellar spacing, which avoid the weak zone at the boundary of austenite and improve the ductility. Besides, the surface hardness of both two
G. Li et al. / Wear 342-343 (2015) 349–355
351
Fig. 1. The geometry of specimen: (a) wheel; (b) rail.
Fig. 2. Microstructure of original steel and specimen A and B: (a) original steel; (b) specimen A, tempered pearlitic steel; (c) specimen B, water sprayed pearlitic steel.
Table 2 Mechanical properties of three steels. Steels
Original Quenched-tempered (spheroidized steel) Water-sprayed cooling (lamellar steel)
Table 3 Experiment parameters of rolling contact fatigue.
Ultimate tensile strength
Yield tensile Elongation Hardness strength
(MPa)
(MPa)
(%)
(HRC)
849.7 955.6
545.2 871.6
18.9 15.2
26.7 32
903.5
667.1
17.6
32
steels was controlled to 32 HRC, which eliminated the influence of surface hardness to fatigue behavior. 3.2. Rolling contact fatigue test Due to the great improvements of mechanical properties after the heat treatment, the author of this work focused on the fatigue behaviors of heat-treated steels and the fatigue tests were only conducted on those two heat-treated steels. According to the YB/T 5345-2006, once the area of deep flaking exceeds 3 mm2 or more, the materials are considered to be failure. However, if the wheels maintain its properties after 1*107 stress cycles, the materials are considered to be fatigue-free and experiments can be ended. The parameters and results of rolling contact fatigue tests are displayed in Tables 3 and 4. It can be concluded that both two kinds of pearlitic steels presented excellent rolling contact fatigue resistance under relatively low speed and pressure. However, lamellar pearlitic steels demonstrated better fatigue resistance under higher speed (2800 rpm) and higher pressure (1200 MPa). Since the surface and cross-section morphology of both two kinds
Group
Creepage (%)
Rotation speed (rpm)
Normal pressure (MPa)
1 2 3
1.6 1.6 1.6
2200 2200 2800
1000 1200 1000
of steels is similar, the following description and discussion are going to focus on the specimen in group 3. During the fatigue test, the quenched-tempered pearlitic steel (specimen A) presented two deep pits after 4.23*106 stress cycles and the areas of two pits were 2 mm2, while the water-sprayed cooling pearlitic steel remained almost complete after 107 stress cycles (Fig. 3). Fig. 4 demonstrates the weight loss after each 5*105 stress cycles. It can be concluded that the weight loss of quenchedtempered steel was always larger than water-sprayed steel. Besides, the weight loss rate of water-sprayed steel remained nearly unchanged, while the weight loss rate of quenchedtempered steel kept rising during the test. 3.3. Worn surface of the wheel Fig. 5 demonstrates the worn surface of the specimens after fatigue experiments. The worn surface of quenched-tempered steel was more complete but the holes on the surface were much deeper than the water-sprayed steel. The worn surface of water-sprayed cooling steel, where presented a number of shallow flat hollows, suggested that lots of flake debris had been removed from the surface. Besides those holes, there existed a lot of scratches on the surfaces of both two steels, which was certainly
352
G. Li et al. / Wear 342-343 (2015) 349–355
induced by hard particles on counterfeit rails or the third body between the rail and wheel. 3.4. Profile of the specimen after rolling contact fatigue test Fig. 6 displays the typical profiles of the specimens after the rolling contact fatigue test. It can be seen from Fig. 6(a) that there lay a number of dark dots below the surface of quenchedtempered steel. Most of these dark dots were located within 20 μm deep. Since the dark regions under electronic microscope represent the lower region of the profile after erosion, these dark Table 4 Results of rolling contact fatigue tests. Group Material
Lifespan 105 cycles
Description
1
1000
No fatigue was observed
2
3
Quenched-tempered steel Water-sprayed steel Quenched-tempered steel Water-sprayed steel Quenched-tempered steel Water-sprayed steel
1000 568 1000 423 1000
Fatigue pits presented No fatigue was observed Fatigue pits presented No fatigue was observed
dots might be the dislocation pile-up groups. Because of the stress introduced during the fatigue test, dislocations began to glide or climb and finally piled up at the obstacles. Those dislocation pileup groups would be easily eroded and leave a dark region at the SEM image. However, little such dark dots existed at a deeper area. Profile of water-sprayed steed presented by Fig. 6(b) displays some different features. It is clear that the lamellar pearlite within 5 μm under the surface had been turned to be parallel with the surface. This direction is also the sliding direction. In this rotating area, some cracks and flaws were observed. Besides, the closer to the surface, the smaller the distance of the lamellar spacing of the pearlite would be. This phenomenon is much clearer in Fig. 7, which demonstrates the comparison of lamellar spacing at 30 μm below the surface of the tested specimen and original specimen. It can be concluded that lamellar spacing decreased from 0.3 μm to 0.15 μm at the 10 μm below the contact surface. 3.5. Micro-hardness results The micro hardness results are demonstrated in Fig. 8. Since the original hardness of both two steels is 32 HRC, which is corresponding to 319 HV in Vicker's hardness, it can be concluded that the working hardening zone of quenched-tempered pearlitic steel is around 150 μm below the surface, while that of water-sprayed cooling pearlitic steel is 300 μm approximately. From Fig. 7, it can also be drawn that the magnitude of working hardening of lamellar pearlitic steel is larger than spheroidized pearlitic steel. This
Fig. 3. Macromorphology of two pearlitic steels after fatigue test: (a) quenched-tempered steel and (b) water-sprayed steel.
G. Li et al. / Wear 342-343 (2015) 349–355
conclusion is also proved by the surface hardness of two steels. The surface hardness of worn spheroidized pearlitic and lamellar pearlitic steel is 356.4 and 395.3 HV respectively.
4. Discussion Earlier paper [5] concluded that most of the cracks initiated at the surface of the wheel or at the sub-surface of the wheel, where the maximum shear stress occurred. It also suggested that the maximum shear stress often occurred at around 4–5 mm below the surface. In this research, no cracks were observed at that area of both two steels. In fact, no cracks were detected at
Fig. 4. Weight losses of two specimens.
353
1 mm or more below the surface. Therefore, fatigue behavior and service life of wheel were mainly determined by the surface initiated cracks. From the experimental data of rolling contact fatigue test, lamellar pearlitic steel demonstrated a better fatigue resistance than spheroidized pearlitic steel. Such behavior may attributed to the different interaction between wear and fatigue for those two different steels. The interaction between wear and fatigue, also known as tribo-fatigue [15,16], proposed that the damage caused by wear and fatigue cannot be simply added when the materials were undertaking both wear and fatigue. Instead, the wear may reduce the fatigue damage. Some researchers [11,12,14,17] have proposed that lamellar pearlitic steel with small lamellar spacing demonstrates great wear resistance because thin cement sheet show better ductility than spheroidized cement. In this research, the plastic deformation area of lamellar pearlitic steel is much larger than the spheroidized pearlitic steel. Thin lamellar perlite with better ductility, together with larger plastic deformation area, would certainly relieve more stress concentration and dissipate more energy. Naturally, with more energy dissipated, the propagation speed of cracks in lamellar pearlitic steel would be slower. On the other hand, the different worn surface morphology and distinguishing mechanical properties of two different pearlitic steels suggest totally different tribo-fatigue behaviors. Although the thin lamellar structure improves the ductility of pearlite, such structure also increases the magnitude of working hardening of ferrite lamella [14,17]. The thin ferrite lamella experienced larger strain than the ferrite matrix in spheroidized pearlite and presented larger working hardening (Fig. 8). Thus, ductile exhaustion of thin ferrite lamella might be induced by such severe plastic deformation and the ferrite region became the weak region
Fig. 5. Worn surface of two specimens: (a) quenched-tempered steel; (b) water-sprayed steel.
Fig. 6. Typical profiles of the specimens after fatigue test: (a)quenched-tempered steel; (b) water-sprayed steel.
354
G. Li et al. / Wear 342-343 (2015) 349–355
Fig. 7. Lamellar spacing of original water-sprayed steel (a) and steel after fatigue test (b).
Fig. 8. Micro-hardness of two steels against distance below the surface.
eventually attenuates the fatigue damage and reduces the weight loss of materials. Instead of this above-mentioned crack propagation process in lamellar steel, the propagation of cracks in spheroidized steel presented a typical propagation process of fatigue cracks. The cracks grew into the material at a shallow angle and then deviated to the sliding direction of the wheel. Then, the crack deviated again to the surface of material and left a deep hole on the surface of the wheel. Fig. 9 demonstrates a crack that almost grew to the surface again. Such a propagation process was occurred at much deeper area below the surface; thus the wear process had little influence on this fatigue process.
5. Conclusions Fig. 9. Crack propagation of spheroidized steel.
of materials. When the surface initiated crack propagated to this weak region, the crack may deviated to this region and propagated along this lamellar region. Once the crack intersected with another surface initiated crack, a flake of debris was removed from the surface. Since the near-surface area (0–5 μm below the surface) had suffered a large amount of plastic deformation, as presented in Fig. 5 (those defects in this region), the above-mentioned propagation of cracks might occurred in this region with a high possibility. The flat shallow pits on the surface give evidence of this crack propagation process. However, this propagation process occurred at the place that extremely close to the surface, where wear damage may also happened. Once the region was removed by wear damage, such fatigue damage was “eliminated” by wear damage. Such competitive rather than superimposed relationship
Two kinds of pearlitic steels with different microstructures were achieved by distinct heat treatment in this work. Tensile strength of both two steels increased a lot while the elongation of materials decreased a little after the heat treatment. The spheroidized steel presented a higher strength but the surface hardness for both two steels was maintained at the same value of 32 HRC. Rolling contact fatigue tests of both two steels suggested that the spheroidized steel failed at 4.23*106 cycles with two deep fatigue pits of around 2 mm2, while the lamellar steel allowed no obvious pits after 107 cycles. Meanwhile, weight loss of spheroidized steel is larger than lamellar steel during the fatigue tests. Those results demonstrated that lamellar steel has better fatigue resistance. The worn surface examination of both two steels displayed that the spheroidized steel maintained a more complete surface with some deep pits while the lamellar one presented a more fragmentary morphology but only with some shallow flake pits. Morphology of
G. Li et al. / Wear 342-343 (2015) 349–355
profile for lamellar steel indicates that the surface initiated cracks propagate along the direction of rolling in the near-surface region (less than 5 μm). Such a special propagation process was induced by the plastic exhaustion of lamellar ferrite and would lead to an unusual tribo-fatigue behavior. The wear on the surface of materials may eliminate the fatigue cracks in the near-surface and thus elevate the fatigue resistance of lamellar steel. Moreover, microhardness tests of the experimented specimens suggest that lamellar steel holds a larger plastic deformation zone. Thus more energy would be delivered to plastic deformation and naturally, less energy would be delivered to cracks propagation. Combining those two aspects, lamellar steel presents better fatigue resistance. Tribo-fatigue view in this paper provides a novel supplement for RCF problem and the results of this research suggest a promising future for lamellar steel equipped as wheel material for high-speed-railway.
Acknowledgments This work was supported by a grant from the Major State Basic Research Development Program of China (973 Program) (No. 2011CB711103).
References [1] A.K. Jha, B.K. Prasad, O.P. Modi, S. Das, A.H. Yegneswaran, Correlating microstructural features and mechanical properties with abrasion resistance of a high strength low alloy steel, Wear 254 (1–2) (2003) 120–128. [2] L. Deters, M. Proksch, Friction and wear testing of rail and wheel material, Wear 258 (7) (2005) 981–991. [3] J.H. Beynon, J.E. Garnham, K.J. Sawley, Rolling contact fatigue of three pearlitic rail steels, Wear 192 (1–2) (1996) 94–111. [4] J.E. Garnham, C.L. Davis, Very early stage rolling contact fatigue crack growth in pearlitic rail steels, Wear 271 (1–2) (2011) 100–112.
355
[5] A. Ekberg, E. Kabo, Fatigue of railway wheels and rails under rolling contact and thermal loading—an overview, Wear 258 (7–8) (2005) 1288–1300. [6] D. Benuzzi, E. Bormetti, G. Donzella, Stress intensity factor range and propagation mode of surface cracks under rolling–sliding contact, Theor. Appl. Fract. Mech. 40 (1) (2003) 55–74. [7] D.F. Cannon, H. Pradier, Rail rolling contact fatigue Research by the European Rail Research Institute, Wear 191 (1–2) (1996) 1–13. [8] D. Nelias, C. Jacq, G. Lormand, G. Dudragne, A. Vincent, New methodology to evaluate the rolling contact fatigue performance of bearing steels with surface dents: application to 32CrMoV13 (nitrided) and M50 steels, J. Tribol. 127 (3) (2005) 611–622. [9] D.I. Fletcher, P. Hyde, A. Kapoor, Modelling and full-scale trials to investigate fluid pressurisation of rolling contact fatigue cracks, Wear 265 (9–10) (2008) 1317–1324. [10] J.W. Ringsberg, Life prediction of rolling contact fatigue crack initiation, Int. J. Fatigue 23 (7) (2001) 575–586. [11] P. Clayton, The relations between wear behaviour and basic material properties for pearlitic steels, Wear 60 (1) (1980) 75–93. [12] J. Kalousek, D.M. Fegredo, E.E. Laufer, The wear resistance and worn metallography of pearlite, bainite and tempered martensite rail steel microstructures of high hardness, Wear 105 (3) (1985) 199–222. [13] R. Kuziak, T. Zygmunt, A new method of rail head hardening of standard-gauge rails for improved wear and damage resistance, Steel Res. Int. 84 (1) (2013) 13–19. [14] Y. Wang, T. Lei, J. Liu, Tribo-metallographic behavior of high carbon steels in dry sliding: II. Microstructure and wear, Wear 231 (1) (1999) 12–19. [15] L.A. Sosnovskiy, Tribo-Fatigue: Wear-Fatigue Damage and its Prediction, Springer Science & Business Media, Berlin, Heidelberg, 2010. [16] L.A. Sosnovskiy, Surprises of tribo-fatigue, in: Proceedings of the World Tribology Congress III, American Society of Mechanical Engineers, 2005. [17] J.E. Garnham, C.L. Davis, The role of deformed rail microstructure on rolling contact fatigue initiation, Wear 265 (9–10) (2008) 1363–1372.