- Email: [email protected]
A novel laser surface compositing by selective laser quenching to enhance railway service life
Author’s Accepted Manuscript A Novel Laser Surface Compositing by Selective Laser Quenching to Enhance Railway Service Life Yinlan Zheng, Qianwu Hu, Chongyang Li, Dengzhi Wang, Li Meng, Jianguo Luo, Juping Wang, Xiaoyan Zeng www.elsevier.com/locate/jtri
PII: DOI: Reference:
S0301-679X(16)30330-9 http://dx.doi.org/10.1016/j.triboint.2016.09.020 JTRI4367
To appear in: Tribiology International Received date: 15 August 2016 Accepted date: 12 September 2016 Cite this article as: Yinlan Zheng, Qianwu Hu, Chongyang Li, Dengzhi Wang, Li Meng, Jianguo Luo, Juping Wang and Xiaoyan Zeng, A Novel Laser Surface Compositing by Selective Laser Quenching to Enhance Railway Service Life, Tribiology International, http://dx.doi.org/10.1016/j.triboint.2016.09.020 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A Novel Laser Surface Compositing by Selective Laser Quenching to Enhance Railway Service Life Yinlan Zhenga,b, Qianwu Hub*, Chongyang Lia, Dengzhi Wangb, Li Mengb, Jianguo Luoc, Juping Wangc, Xiaoyan Zenga,b* a
School of Optical and Electronic Information Wuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan 430074, Hubei, P. R. China. c Institute of Railway Technique, Wuhan Railway Ltd., Wuhan 430074, Hubei, P. R. China. b
[email protected] [email protected] *
Corresponding author. Tel/Fax: +86 27 87541423.
Abstract With the rapid development of high-speed and heavy-haul trains, it is challenging to improve the service life of rails. In order to ensure toughness and contact fatigue resistance (CFR) of rails, the present technical standards restrict the hardness of rails to be beneath HB400 and forbid the martensite structure in rails. In this study, a novel laser surface compositing (LSC) based on selective laser quenching is proposed to strengthen the rails. By designing the hardening area sizes and proportions rationally, the wear resistance of the rails is enhanced by 2.2-3.5 times without noticeably deteriorating their contact fatigue resistance (CFR). Furthermore, the wear and contact fatigue behaviours of the LSC rails were investigated systematically. Keywords: Laser surface compositing (LSC); Selective laser quenching; Wear resistance; Contact fatigue resistance (CFR).
1. Introduction With the rapid development of high-speed and heavy-haul trains all over the world, the short service life of rails, especially for some key parts of the railways, has become a serious challenge. This affects the safety and transportation capacity of trains significantly.
1
When the train runs on the rail, the contact area for each wheel/rail is only about 100 mm2 and endures dozens of tons of alternating vertical loads. This easily leads to wear, spalling, compression, and even fracture of the rails, especially for turnouts, switch rails, and sharp curve rails in heavy load railways. It is well known that both the wear resistance and the contact fatigue resistance (CFR) are two key, but contradictory, factors that determine the service life of rails [1, 2]
. Because of the complicated alternating load, the fatigue cracks are prone to
nucleate and propagate in martensite structures, which threaten the safe transportation of the trains. Therefore, the martensite structure has been forbidden in the technical standards worldwide. Almost all of these standards also include a rail hardness below HB400 [3]. For example, the microstructure of rail steels is usually pearlitic with a small amount of ferrite and no martensite, and this is permitted by the American Railway Engineering and Maintenance-of-Way Association Standard, the draft European Rail Standard, and the Chinese Railway Standard [4-6]. In order to enhance the hardness of the rails while also keeping the contact fatigue resistance at high levels, extensive studies have been carried out by developing new rail materials, such as bainite steel, and by surface engineering technologies, such as induction quenching and immediate quenching treatment
[7.8]
.
However, to the authors’ knowledge, no apparent breakthrough has been found in the past decades. Laser quenching is a surface hardening technique by irradiating the steel surface with a focused laser beam, thus generating martensite transformation in site after the beam moves away. It is a kind of clean and fast quenching process with the advantages of high power density, fast cooling speed, and no cooling media such as water or oil. Compared with the induction quenching and carburizing quenching process, laser quenching has a uniform hardening layer, high hardness, small deformation, and easy control of the hardening depth and tracks. Several years ago, the U.S. Department of Transportation assessed the effect of laser glazing on rail surfaces to reduce the use of lubricants and enhance the service 2
life, but the fatigue problems were too hard to resolve. Their results suggested that the cracks formed during the laser treatment process led to the early spalling and chipping of the gage corner [9]. In this study, a novel method named laser surface compositing (LSC) based on selective laser quenching is put forward to strengthen the rail surface. By controlling the LSC area appropriately, martensite hardening arrays with hardness of HV800 are distributed uniformly in the toughness pearlite rail matrix. Due to the synergistic effect of the martensite hardening arrays and the toughness pearlite rail matrix, the LSC enhanced wear resistance and service life of rails by 2.2-3.5 times. Furthermore, the wear and contact fatigue behaviours of the LSC rails are analysed.
2. Materials and Methods for Wear and CFR Tests 2.1 Materials The rail and wheel material is U71Mn and CL60, respectively. Their chemical compositions are shown in Table 1. To prepare the specimens, the raw materials of U71Mn and CL60 were cut from rail head and wheel rim in order to guarantee the conformance of materials and structures with real rail/wheel. Table1.Chemical composition of U71Mn and CL60 (wt.%). Material
C
Si
Mn
P
S
V
Nb
Cr
Ni
U71Mn
0.65~0.76
0.15~0.35
1.10~1.40
≤0.03
≤0.03
≤0.03
≤0.01
/
/
CL60
0.57~0.65
0.17~0.37
0.50~0.80
≤0.035
≤0.035
/
/
≤0.25
≤0.25
2.2 Specimen Preparations LSC is carried out by a fiber laser (YLR-6000, IPG) with maximum power of 6kW and a wavelength of 1.07 μm. The laser was transmitted by a 200 μm diameter core fiber and focused by a beam shaping lens to get a flattened laser intensity profile. In order to investigate the influences of the selective hardening area on wear and CFR of the rail, 5 groups of specimens with different hardening area ratios of 100%, 3
49.8%, 48.6%, 28%, and 0% (Fig. 1) are treated by LSC. For LSC specimens, the hardening area is the same as the hardening spot size all together. The hardening area ratio is defined as the ratio of the total hardening spot area to the whole surface area. The detailed experimental parameters are listed in Table 2.
Figure1. Parameters of 5 specimens treated by LSC with different hardening area ratios: 0% (#0), 100% (#1), 49.8% (#2), 48.6% (#3), 28% (#4). Table2. The rail specimens and corresponding laser quenching parameters. Specimen number
#0
#1
#2
#3
#4
Hardening area
--
All-area
LSC
LSC
LSC
Laser power (W)
--
1200
550
350
200
Laser spot diameter (mm)
--
4
3
2
1
Irradiating time (s)
--
--
0.3
0.3
0.3
Distance of dot (mm)
--
--
4.1
2.7
1.8
Scanning speed (mm/s)
--
16.7
--
--
--
Hardening area ratio
0
100%
49.8%
48.6%
28%
Hardening depth(mm)
0
0.74
0.64
0.45
0.095
2.3 Wear and CFR Tests To ensure the compatibility of the experimental results with the real working condition, a roller-on-roller model was selected to simulate the contact between the wheel and the rail treads. A special testing apparatus, named the contact fatigue simulation testing apparatus (WTY-1), was used to test wear resistance and contact fatigue resistance of the locomotive railway under the straight trajectory running conditions. The testing apparatus is composed of a hydraulic pressure loading part, an electric motor, a gear driving part, and so on, as shown in Fig. 2. Axial loading is applied by the hydraulic system, and then the straight trajectory working condition of 60 kg/m railway is simulated. The testing apparatus drawing is in Fig. 3. The whole system is designed to simulate the contact between the wheel flange and rail tread.
4
The wheel and rail specimen drawings and the contact state of specimens is in Fig. 4. The design of specimen geometries and the test equipment contribute to establish a uniform contact between the specimens and to control the contact pressure. The rail specimens are cylindrical discs, while the wheel specimens use a cylindrical disc shape with a fillet in order to allow high contact pressure tests under reasonable loads.
Figure 2. Front view (a) and side view (b) of the wear resistance and CFR simulation testing apparatus used in the experiment (WTY-1).
Figure 3. The testing apparatus drawing.
5
Figure 4. The wheel and rail specimen drawings and the contact state of specimens for CFR tests.
A typical field running condition of straight trajectory of 60 kg/m railway is chosen to be simulated, and the detailed parameters of the field running conditions are as follows: the diameter of the wheel is 0.625 m, the curvature radius of the wheel is 0.3 m, the axial loads are 2.5×104 kg, and the running speed is 61.1 m/s. Rolling-sliding tests were conducted to investigate the condition above, and all the tests were conducted with a 2% slip (“slip” means the difference in rolling speed between the rail and the wheel specimen). The testing parameters are calculated according to the field running conditions based on Hertz Contact Theory [10], as shown in Table 3. According to Fig. 4, the semi-major axis is one half of the major axis of the contact ellipse, and the semi-minor axis is one half of the shorter axis of the contact ellipse. In order to prevent wheel specimens from wearing excessively during the test, the wheels were hardened to HRC50-55 by high-frequency induction quenching. Table 3. Rolling contact parameters for wear resistance and CFR experiment. Field axial loads (kg)
2.5×104
Simulation axial loads (kg)
120
Semi-major axis (m)
8.78×10-4
Semi-minor axis (m)
4.54×10-4
Maximum contact stress (MPa)
1400
Slip
2%
Experiment cycles
250,000
3. Results 6
3.1 Microstructure and Microhardness The cross-section morphologies of the hardening layer treated by LSC are shown in Fig. 5. The hardening layer of #1 (all area treated by LSC) is flat and continuously uniformly distributed on the rail surface, while the hardening layers of the other specimens of #2, #3, and #4 (selectively treated by LSC) are crescent-shaped and isolated.
Figure 5. The cross-section morphologies of the hardening layer.
Figure 6 shows the cross-section microstructure of specimen #2 treated by LSC. The structure was characterized of hardening layer, heat affected zone (HAZ), and substrate, as shown in Fig. 6(a). In Fig. 6(b), small acicular martensite is found in the hardening layer and was produced by rapid heating and cooling of laser quenching, while the growth of austenite was suppressed. The HAZ is composed of martensite, residual austenite, and partial substrate，as shown in Fig. 6(c). The substrate of the rail is composed of typical lamellar pearlite, as shown in Fig. 6(d). Similar structures can be found in specimen #3 and #4.
7
Figure 6. (a) Structure of specimen #2 from surface to substrate; SEM of (b) the hardening layer, (c) the HAZ and (d) the substrate.
After experiment, all the specimens were cut at the middle of the hardening spots, and microhardness distribution was tested along their depth direction, as shown in Fig. 7. The hardness of specimen #1~#4 by LSC is in the range of HV750~850. The hardness of the HAZ is HV400~HV500. The average hardness of the untreated area of the substrate is about HV300.
Figure 7. Microhardness distribution along depth direction of the specimens.
3.2 Wear
8
The weight loss was estimated by weighing specimens before and after the tests. The wear losses of the specimens were measured by an electronic balance (Mettler-Toledo AL204/1) with an accuracy of 0.1 mg. Relative wear resistance R for different specimens is defined by Formula (1) [12]： R=Wb/Wa
(1)
where Wb is the wear losses of the untreated rail specimen after the wear and CFR tests, and Wa is the wear losses of the LSC-treated rail specimen after wear test. Figure 8 shows the wear resistance of the specimens. It is found that the relative wear resistance of specimens treated by LSC is 1.2~40 times higher than that of the untreated specimens, and the relative wear resistance of specimens #1~#4 is 40.3, 9.7, 2.5 and 1.2, respectively, decreasing with the reduction of quenched depth and hardening area ratio.
Figure 8. Wear resistance of the specimens under different processing conditions.
Figure 9(a) shows the surface morphologies of the wear specimens. The widths of the wear patterns on the surfaces of specimens #0 and #4 are much wider than those on specimens #1~#3, as marked by the yellow arrows. This means that the surface plastic deformation of the rail materials is severe. The side wave deformation, pointed out by the blue arrows in specimens #0 and #4, also indicates that both specimens have endured severe deformation. The main reason may be due to that specimen #0 has a low hardness level and specimen #4 has shallow hardening layers and a smaller hardening area ratio.
9
On the contrary, the side wave deformation of specimens #1~#3 with high hardening area ratios is slight or even invisible. It should be noted that not all the LSC-treated specimens have excellent wear resistance. For example, the surface morphology of specimen #4 is similar to the untreated specimen #0. On the other hand, although specimens #2 and #3 have excellent wear resistance, part of the hardening layers under the contact areas are worn away, as shown by the red dashed circles and the residual layers pointed out by black arrows. Some small cracks and spalling can be observed on the wear surface, indicated by the purple arrows. Profilometry was performed to assess the wear depth on the rail specimens (Fig. 9(b)) and to compare and validate the results. On the surface of specimen #0 without laser quenching, an extended groove and deep wear pattern (0.57mm) demonstrate that the wear between the rail and wheel is very serious, which would affect the service life of the rails. In contrast, the wear depth of specimen #1 is very thin. Compared with specimen #0, the wear resistance by LSC treatment (specimens #2~#4) can be improved, but the extent depends on the hardening depths and the area ratios of the quenched dots. The larger the ratios and depths are, the better the wear resistance of the specimens treated by LSC is.
Figure 9. (a) Surface morphologies of specimens after experiment; (b) wear depth of specimens after experiment.
3.3 Contact Fatigue Resistance (CFR) 10
In this research, the depth and length of cracks were analysed to evaluate the CFR of rails. The cross-section morphologies along the longitudinal direction of specimen #0 are shown in Figs. 10(a)-(b), in which there are several cracks with lengths about 10 μm and depths smaller than 50 μm. These shallow and short cracks are prone to be removed by the wear, and no penetrating crack is found in specimen #0. Hence, the service life of the untreated rail is mainly determined by its wear resistance. The cross section of specimen #1 is shown in Figs. 10(c)-(e), in which neither plastic deformation nor a micro-crack can be observed due to its high hardness. However, a crack nucleates and propagates from the hardening layer to the substrate. The leading crack ramifies into several secondary cracks in different directions in the interface between the hardening layer and the substrate, as shown in Fig. 10(c). The depth of the leading crack is 1000 μm, which is 17 times as much as the wear depth (58μm). Further, the downward propagation branch shown in Fig. 10(e) may cause rail damage or fracture during long-term service, which may threaten the transportation safety of the trains. Thus, it should be emphasized that it is the CFR, rather than the wear resistance, that determines the service life of laser all-area hardened rails. This result can explain why whole-surface hardening, either by laser quenching or other traditional quenching techniques, is forbidden in rail strengthening [4-6]
.
11
Figure 10. (a, b) Cracks of untreated specimen #0; (c) crack morphology of specimen #1; (d) microstructure of a leading crack of specimen #1; (e) downward crack propagation of specimen #1; (f) cracks and spalling of specimen #3; (g,h) superficial and horizontal crack of specimen #3; (i,j) cracks of specimen #4.
The cross-section morphology of specimen #3 by LSC is shown in Fig. 10(f), in which the contact fatigue cracks nucleate near the boundary between the hardening layer and the substrate, and then propagate along the surface direction with a length of about 0.1 mm, as shown in Figs. 10(g) and 10(h).
12
Compared with specimen #1, the cracks of specimen #3 are shorter, and their positions are shallower. In Fig. 10(h), the cracks in specimen #3 are about 100 μm and are located on the surface with 38% of the wear depth. Most of the cracks in specimen #3 propagate horizontally, and no penetrating crack is found in the hardening layer or substrate. Similar characteristics are observed in the cross-section morphologies of the specimen #2. Figures 10(i)-(j) display the cross-section morphologies along the longitudinal direction of specimen #4, and it can be observed obviously that specimen #4 has similar plastic deformation and wear resistance to specimen #0. It may be because the hardening spot and depth of specimen #4 are too small, and the surface property is almost the same as the as-received state.
4. Discussion Given the amount of data collected in experiment and in field (section 5), it is necessary to declare in this section that our approach is a preliminary and qualitative interpretation of the results to be confirmed by further detailed analyses, instead of carrying out complex calculations at this stage. Generally speaking, contact fatigue cracks in the rail are caused by surface distortion and relaxation under alternating stresses between the wheel and rail. The fatigue failure is divided into two stages: firstly, the crack nucleates and propagates in the rail
[11-13]
(the simplified two-dimensional plane stress model, Hertz Contact, can
explain the cracking behaviour in the travelling direction
[10]
); secondly, the normal
stress PZ(x) and the tangential stress q(x) from friction are applied on the rail. Given the friction coefficient and the normal stress, q(x) is calculated according to Formula (2) and (3) [14]: Pz(x)=PZMAX[1-(x/b)2]1/2 q(x)=μ·PZ(x)
(2) (3)
where PZMAX is the maximum normal stress, b is the semi axis of the contact area for the wheel and rail, and μ is the friction coefficient between the wheel and rail. The 13
induced stress distribution against the normalized distance from the surface is plotted in Fig. 11(a).
Figure 11. Stress analysis of the rail due to the interaction between the rail and wheel: (a) stress distribution on rail; (b) change of maximum shear stress with different friction coefficient μ [14].
Assuming that both the rail and the wheel have perfectly smooth surfaces and no friction occurs between them, only the normal stress from the wheel exists on the rail. Thus, the maximum shear stress (τmax) on the rail surface can be defined by Formula (4) [11]: τmax={[1/2(σx-σz)]2+τxz}1/2
(4)
According to Formula (4), the shear stress isoline can be drawn, and τmax is found in the sub-surface
[11]
. Here, plastic deformation is prone to take place, and contact
fatigue cracks nucleate. This means that the internal shear stress exists in the rail even on an ideal smooth surface, and the crack nucleation is not in the upper surface but in the sub-surface of the rail where τmax is. In realistic engineering conditions, there is no pure rolling for the wheel and rail 14
pair, and the tangential stress caused by wheel/rail friction cannot be ignored. Multiple studies indicate that τmax increases and depth decreases with an increase of the friction coefficient μ
[14-16]
. Figure 11(b) presents the shear stress distributions
under different friction values μ by computational simulation
[14]
. When the friction
coefficient μ changes from 0 to 0.4, τmax increases from 380 MPa to 450 MPa; the position of the cracks will move from sub-surface (i.e. 1 normalized depth) to the upper surface of the rail [14]. Furthermore, the rails are subjected to the mixed-mode cyclic loading, including mode I stress (caused by the normal stress) and mode II stress (caused by the tangential stress) in real engineering conditions. Since the stress mode II tends to change the original crack propagation directions, the wear and the rolling contact fatigue process will become more complicated
[17]
. The stress intensity factors (SIFs)
for mode I and mode II are combined respectively for an effective stress intensity factor to describe crack propagation rates. Tanaka
[16]
and Weertman
[18]
express a
crack propagation rate, similar to Paris’ Law [11], by Formula (5): da/dN=C(ΔKeff)4, ΔKeff=KI+8KII
(5)
where C is the material constant, ΔKeff is the effective SIF, KI and KII are the SIFs of mode I and mode II, respectively, and da/dN represents the crack length increment per loading cycle. Based on Formula (4) and (5), we can discuss the mechanisms of the main physical factors on contact fatigue resistance and the service life of rails. Apart from the material strength of rails, the nucleation and propagation rates of the contact fatigue cracks are determined by the following two conditions: (1) The magnitude and the depth of τmax. The magnitude of τmax determines the crack nucleation time, and the depth of τmax determines the crack nucleation position. The larger the μ or PZ(x) is, the shallower the crack nucleation position would be, and the quicker the crack nucleation would start up.
15
(2) The magnitude of ΔKeff. The value of ΔKeff determines the propagation rates of the contact fatigue cracks after nucleating, and the crack propagating rate is in proportion with the value of ΔKeff. Now we can explain why the laser all-area hardened rails have low fatigue resistance. In typical rail service conditions, as shown in Fig. 12(a), laser all-area quenching will reduce the friction coefficient
[17]
, and make the site of the maximum
shear stress τmax1 deeper, i.e., the crack nucleation location moves inward from the rail’s upper surface
[19]
. Therefore, once the cracks nucleate from the sub-surface, the
cracks cannot be removed by wear and are prone to propagate into long cracks under alternate loadings, greatly shortening the contact fatigue life as well as the service life of the rails. This can explain why the martensite structure of rails is forbidden worldwide. However, laser surface compositing (LSC) has formed a composite surface with the martensite hardening arrays embedded in the untreated pearlite structure matrix. During service, the ductile pearlite is worn away prior by the wheel, and the hardened martensite dot arrays gradually become prominent on the surface, withstanding the main load from the wheel (shown in Fig. 12(b)). The reduction of the contact areas causes the normal and tangential stress to increase greatly. Under this condition, the τmax at the hardening edge (noted as τmax1) gets bigger, and its position (D2) moves upward, meaning the fatigue cracks would probably nucleate at the boundary between the martensite and the matrix on the rail surface, as shown in Fig. 12(c). When the wheels roll at the middle part of the hardening areas, the contact area increases, and the normal stress τmax2 decreases (τmax2＜τmax1). Since the values of KI2 and KII2 in the middle are smaller than those in the boundary areas, the effective SIF at the middle is also smaller than that at the boundary (ΔKeff2＜ΔKeff1). According to Formula (6), (da/dN)2＜(da/dN)1, i.e., the crack propagation rates at the boundary is much faster than those at the middle. As the nucleating cracks at boundaries are superficial, they are prone to be removed by the wear or spalling effect due to the alternating stress between the wheel/rail. 16
Figure 12. Interactions of wheel/rail under different treatment: (a) crack propagation sketch for laser all-area hardened rail; (b) the early stage of experiment of LSC rail; (c) crack propagation sketch for LSC rail.
Summing up the above analysis, the crack propagation mechanisms of the rails treated by LSC are totally different from that by laser all-area quenched rails, and the martensite structure is no longer the dangerous phases for the rails treated by LSC. The rail service life of LSC is dependent upon the CFR and wear resistance, which can be adjusted and controlled by the area proportion and the depths of the hardening arrays. Generally speaking, the CFR and wear resistance are two interactional and competing factors that determine the service life of rails simultaneously
[20]
. On one
hand, the specimens without laser quenching have good toughness and CFR, but its wear resistance is weak due to its low hardness. On the other hand, laser all-area quenched rails have the highest wear resistance. However, the contact fatigue cracks of the samples nucleate easily in the martensite areas and propagate in big angles 17
(nearly 90°), which is harmful and disrupts the rails. When LSC is utilized, the hardening area ratios and depths change greatly with different processing parameters, which may observably affect the wear resistance and contact fatigue resistance. The specimens with bigger hardening area ratios and deeper hardening layers have better wear resistance but worse fatigue resistance. By designing the hardening area sizes and proportions rationally, the service life of the rails can be enhanced by 2.2-3.5 times without obviously deteriorating their CFR.
5. Application Verification More than 600 groups of rails treated by LSC have been applied in Wuhan Railway Bureau of China, as shown in Fig. 13. Table 4 shows the service life of the railway switches by LSC with different laser processing parameters. The results demonstrate that the service life can be enhanced by 2.2-3.5 times as much as that without laser treatment. The service life of those treated by LSC depends on not only the laser processing parameters, but also the service conditions of the rails. The maximum service life can even be enhanced by 10 times the untreated rails. The detailed results will be published in our next paper [21]. Table 4. The service life by laser surface compositing under different processing parameters. Switch Number
#1
#2
#3
Treatment statement
No treatment
LSC
LSC
Service life enhancement
1.0
2.2
3.5
*The parameters by LSC for railway switch #2 and #3 are corresponding to specimen #2 and #3.
18
Figure 13. The railway switches by laser surface compositing are setup in Wuhan Railway Bureau. (a) The workers are setting up the rail switch. (b) The rail switches after setup.
6. Summary (1) A novel laser surface compositing (LSC) is firstly introduced on the rail surface by selective laser quenching, which produced martensite arrays with a hardness of HV800 on the toughness pearlite rail matrix. (2) Due to the synthetic effect of the high hardness of the martensite arrays and good toughness of the rail matrix, LSC can enhance the service life of the rails by 2.2-3.5 times without significant loss in its contact fatigue resistance, and it has broken through the traditional worldwide restriction that forbids the martensite structure on rail surfaces in the past hundred years.
Acknowledgements This research was financially supported by the national key research and development program 2016YFB1102600, P. R. China. The application verification of this research work was supported by the institute of railway technique of Wuhan Railway Ltd., and the authors are grateful to Prof. Yongfeng Lu for valuable discussions.
References [1] W. Zhong, J.J. Hu, P. Shen, C.Y. Wang, Q.Y. Lius, Experimental investigation between rolling contact fatigue and wear of high-speed and heavy-haul railway and selection of rail material, Wear, 271(2011) 2485-2493. [2] G. Donzella, M. Faccoli, A. Ghidini, A. Mazzù, R. Roberti, The competitive role of wear and RCF in a rail steel, Eng. Fract. Mech., 72(2005) 287-308. [3] D.F. CANNON, K.O. EDEL, S.L. GRASSIE, K. SAWLEY, Rail defects: an overview, Fatigue Fract. Eng. M., 26(2003) 865-886. [4] E. Standard, Railway Applications-Track-Rail, Part 1: Vignole railway rails 46kg/m and above, in: 13674-1 13674-1, 2002. [5] AREMA, Manual for railway Engineering SECTION 2 SPECIFICATION FOR 19
STEEL RAILS, 2010. [6] C. Standard, Technical Specifications for the procurement of 43 kg/m ~ 75 kg/m rails, 2004. [7] K. SAWLEY, K. J, Development of bainitic rail steels with potential resistance to rolling contact fatigue, Fatigue Fract. Eng. M., 26(2003) 1019-1029. [8] Y. Totik, R. Sadeler, H. Altun, M. Gavgali, The effects of induction hardening on wear properties of AISI 4140 steel in dry sliding conditions, Materials and Design, 24(2003) 25-30. [9] R. Reiff, In-Track Demonstration of Laser-Treated Rail to Reduce Friction and Wear, U.S. Department of Transportation, 2007. [10] B. H. Hertz, On the contact of elastic solids,Zeitschrift fur Reine und AngewandteMathematik, 92(1882) 156--171. [11] L.M. Keer, Contact Mechanics (K. L. Johnson), The Society for Industrial and Applied Mathematics, 1987. [12] Q.H.X.Z. Dengzhi Wang, Microstructures and performances of Cr13Ni5Si2 based composite coatings deposited by laser cladding and laser-induction hybrid cladding, J. Alloy. Compd, (2014) 502-508. [13] J.H. Beynon, J.E. Garnham, K.J. Sawley, Rolling contact fatigue of three pearlitic rail steels, Wear, 192(1996) 94-111. [14] J. Seo, S. Kwon, H. Jun, D. Lee, Fatigue crack growth behavior of surface crack in rails, Procedia Engineering, 2(2010) 865-872. [15] B. Trollé, M.C. Baietto, A. Gravouil, S.H. Mai, B. Prabel, 2D fatigue crack propagation in rails taking into account actual plastic stresses, Eng. Fract. Mech., 123(2014) 163-181. [16] K. Tanaka, Y. Nakai, M. Yamashita, Fatigue growth threshold of small cracks, Int. J. Fracture, 17(1981) 519-533. [17] R.J. DiMelfi, P.G. Sanders, Mitigation of subsurface crack propagation in railroad rails by laser surface modification1, Surface and Coatings Technology, 106(1998) 30-43. 20
[18] J. Weertman, Theory of fatigue crack growth based on a BCS crack theory with work hardening, Int. J. Fracture, 9(1973) 125-131. [19] T. Kuminek, K. AnioEk, J. M Yńczak, A numerical analysis of the contact stress distribution and physical modelling of abrasive wear in the tram wheel-frog system, Wear, 328–329(2015) 177-185. [20] D.T. Eadie, D. Elvidge, K. Oldknow, R. Stock, P. Pointner, J. Kalousek, P. Klauser, The effects of top of rail friction modifier on wear and rolling contact fatigue: Full-scale rail–wheel test rig evaluation, analysis and modelling, Wear, 265(2008) 1222-1230. [21] Q.H.Y.Z. Xiaoyan Zeng, Service life under different laser processing parameters, to be submitted to Surface and Coatings Technology.
Highlights
In this study, a novel laser surface compositing (LSC) is firstly introduced on the rail surface by selective laser quenching, which produced martensite arrays with a hardness of HV800 on the toughness pearlite rail matrix.
By designing the hardening area sizes and proportions rationally, the wear resistance of the rails is enhanced by 2.2-3.5 times without noticeably deteriorating their contact fatigue resistance (CFR).
Due to the synthetic effect of the high hardness of the martensite arrays and good toughness of the rail matrix, LSC can enhance the service life of the rails by 2.2-3.5 times, and it has broken through the traditional worldwide restriction that forbids the martensite structure on rail surfaces in the past hundred years.
The mechanisms of LSC to enhance the service life of rails are analyzed systemically, and the rails treated by LSC are taken into application to verify the theoretical research. 21
22
PII: DOI: Reference:
S0301-679X(16)30330-9 http://dx.doi.org/10.1016/j.triboint.2016.09.020 JTRI4367
To appear in: Tribiology International Received date: 15 August 2016 Accepted date: 12 September 2016 Cite this article as: Yinlan Zheng, Qianwu Hu, Chongyang Li, Dengzhi Wang, Li Meng, Jianguo Luo, Juping Wang and Xiaoyan Zeng, A Novel Laser Surface Compositing by Selective Laser Quenching to Enhance Railway Service Life, Tribiology International, http://dx.doi.org/10.1016/j.triboint.2016.09.020 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A Novel Laser Surface Compositing by Selective Laser Quenching to Enhance Railway Service Life Yinlan Zhenga,b, Qianwu Hub*, Chongyang Lia, Dengzhi Wangb, Li Mengb, Jianguo Luoc, Juping Wangc, Xiaoyan Zenga,b* a
School of Optical and Electronic Information Wuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan 430074, Hubei, P. R. China. c Institute of Railway Technique, Wuhan Railway Ltd., Wuhan 430074, Hubei, P. R. China. b
[email protected] [email protected] *
Corresponding author. Tel/Fax: +86 27 87541423.
Abstract With the rapid development of high-speed and heavy-haul trains, it is challenging to improve the service life of rails. In order to ensure toughness and contact fatigue resistance (CFR) of rails, the present technical standards restrict the hardness of rails to be beneath HB400 and forbid the martensite structure in rails. In this study, a novel laser surface compositing (LSC) based on selective laser quenching is proposed to strengthen the rails. By designing the hardening area sizes and proportions rationally, the wear resistance of the rails is enhanced by 2.2-3.5 times without noticeably deteriorating their contact fatigue resistance (CFR). Furthermore, the wear and contact fatigue behaviours of the LSC rails were investigated systematically. Keywords: Laser surface compositing (LSC); Selective laser quenching; Wear resistance; Contact fatigue resistance (CFR).
1. Introduction With the rapid development of high-speed and heavy-haul trains all over the world, the short service life of rails, especially for some key parts of the railways, has become a serious challenge. This affects the safety and transportation capacity of trains significantly.
1
When the train runs on the rail, the contact area for each wheel/rail is only about 100 mm2 and endures dozens of tons of alternating vertical loads. This easily leads to wear, spalling, compression, and even fracture of the rails, especially for turnouts, switch rails, and sharp curve rails in heavy load railways. It is well known that both the wear resistance and the contact fatigue resistance (CFR) are two key, but contradictory, factors that determine the service life of rails [1, 2]
. Because of the complicated alternating load, the fatigue cracks are prone to
nucleate and propagate in martensite structures, which threaten the safe transportation of the trains. Therefore, the martensite structure has been forbidden in the technical standards worldwide. Almost all of these standards also include a rail hardness below HB400 [3]. For example, the microstructure of rail steels is usually pearlitic with a small amount of ferrite and no martensite, and this is permitted by the American Railway Engineering and Maintenance-of-Way Association Standard, the draft European Rail Standard, and the Chinese Railway Standard [4-6]. In order to enhance the hardness of the rails while also keeping the contact fatigue resistance at high levels, extensive studies have been carried out by developing new rail materials, such as bainite steel, and by surface engineering technologies, such as induction quenching and immediate quenching treatment
[7.8]
.
However, to the authors’ knowledge, no apparent breakthrough has been found in the past decades. Laser quenching is a surface hardening technique by irradiating the steel surface with a focused laser beam, thus generating martensite transformation in site after the beam moves away. It is a kind of clean and fast quenching process with the advantages of high power density, fast cooling speed, and no cooling media such as water or oil. Compared with the induction quenching and carburizing quenching process, laser quenching has a uniform hardening layer, high hardness, small deformation, and easy control of the hardening depth and tracks. Several years ago, the U.S. Department of Transportation assessed the effect of laser glazing on rail surfaces to reduce the use of lubricants and enhance the service 2
life, but the fatigue problems were too hard to resolve. Their results suggested that the cracks formed during the laser treatment process led to the early spalling and chipping of the gage corner [9]. In this study, a novel method named laser surface compositing (LSC) based on selective laser quenching is put forward to strengthen the rail surface. By controlling the LSC area appropriately, martensite hardening arrays with hardness of HV800 are distributed uniformly in the toughness pearlite rail matrix. Due to the synergistic effect of the martensite hardening arrays and the toughness pearlite rail matrix, the LSC enhanced wear resistance and service life of rails by 2.2-3.5 times. Furthermore, the wear and contact fatigue behaviours of the LSC rails are analysed.
2. Materials and Methods for Wear and CFR Tests 2.1 Materials The rail and wheel material is U71Mn and CL60, respectively. Their chemical compositions are shown in Table 1. To prepare the specimens, the raw materials of U71Mn and CL60 were cut from rail head and wheel rim in order to guarantee the conformance of materials and structures with real rail/wheel. Table1.Chemical composition of U71Mn and CL60 (wt.%). Material
C
Si
Mn
P
S
V
Nb
Cr
Ni
U71Mn
0.65~0.76
0.15~0.35
1.10~1.40
≤0.03
≤0.03
≤0.03
≤0.01
/
/
CL60
0.57~0.65
0.17~0.37
0.50~0.80
≤0.035
≤0.035
/
/
≤0.25
≤0.25
2.2 Specimen Preparations LSC is carried out by a fiber laser (YLR-6000, IPG) with maximum power of 6kW and a wavelength of 1.07 μm. The laser was transmitted by a 200 μm diameter core fiber and focused by a beam shaping lens to get a flattened laser intensity profile. In order to investigate the influences of the selective hardening area on wear and CFR of the rail, 5 groups of specimens with different hardening area ratios of 100%, 3
49.8%, 48.6%, 28%, and 0% (Fig. 1) are treated by LSC. For LSC specimens, the hardening area is the same as the hardening spot size all together. The hardening area ratio is defined as the ratio of the total hardening spot area to the whole surface area. The detailed experimental parameters are listed in Table 2.
Figure1. Parameters of 5 specimens treated by LSC with different hardening area ratios: 0% (#0), 100% (#1), 49.8% (#2), 48.6% (#3), 28% (#4). Table2. The rail specimens and corresponding laser quenching parameters. Specimen number
#0
#1
#2
#3
#4
Hardening area
--
All-area
LSC
LSC
LSC
Laser power (W)
--
1200
550
350
200
Laser spot diameter (mm)
--
4
3
2
1
Irradiating time (s)
--
--
0.3
0.3
0.3
Distance of dot (mm)
--
--
4.1
2.7
1.8
Scanning speed (mm/s)
--
16.7
--
--
--
Hardening area ratio
0
100%
49.8%
48.6%
28%
Hardening depth(mm)
0
0.74
0.64
0.45
0.095
2.3 Wear and CFR Tests To ensure the compatibility of the experimental results with the real working condition, a roller-on-roller model was selected to simulate the contact between the wheel and the rail treads. A special testing apparatus, named the contact fatigue simulation testing apparatus (WTY-1), was used to test wear resistance and contact fatigue resistance of the locomotive railway under the straight trajectory running conditions. The testing apparatus is composed of a hydraulic pressure loading part, an electric motor, a gear driving part, and so on, as shown in Fig. 2. Axial loading is applied by the hydraulic system, and then the straight trajectory working condition of 60 kg/m railway is simulated. The testing apparatus drawing is in Fig. 3. The whole system is designed to simulate the contact between the wheel flange and rail tread.
4
The wheel and rail specimen drawings and the contact state of specimens is in Fig. 4. The design of specimen geometries and the test equipment contribute to establish a uniform contact between the specimens and to control the contact pressure. The rail specimens are cylindrical discs, while the wheel specimens use a cylindrical disc shape with a fillet in order to allow high contact pressure tests under reasonable loads.
Figure 2. Front view (a) and side view (b) of the wear resistance and CFR simulation testing apparatus used in the experiment (WTY-1).
Figure 3. The testing apparatus drawing.
5
Figure 4. The wheel and rail specimen drawings and the contact state of specimens for CFR tests.
A typical field running condition of straight trajectory of 60 kg/m railway is chosen to be simulated, and the detailed parameters of the field running conditions are as follows: the diameter of the wheel is 0.625 m, the curvature radius of the wheel is 0.3 m, the axial loads are 2.5×104 kg, and the running speed is 61.1 m/s. Rolling-sliding tests were conducted to investigate the condition above, and all the tests were conducted with a 2% slip (“slip” means the difference in rolling speed between the rail and the wheel specimen). The testing parameters are calculated according to the field running conditions based on Hertz Contact Theory [10], as shown in Table 3. According to Fig. 4, the semi-major axis is one half of the major axis of the contact ellipse, and the semi-minor axis is one half of the shorter axis of the contact ellipse. In order to prevent wheel specimens from wearing excessively during the test, the wheels were hardened to HRC50-55 by high-frequency induction quenching. Table 3. Rolling contact parameters for wear resistance and CFR experiment. Field axial loads (kg)
2.5×104
Simulation axial loads (kg)
120
Semi-major axis (m)
8.78×10-4
Semi-minor axis (m)
4.54×10-4
Maximum contact stress (MPa)
1400
Slip
2%
Experiment cycles
250,000
3. Results 6
3.1 Microstructure and Microhardness The cross-section morphologies of the hardening layer treated by LSC are shown in Fig. 5. The hardening layer of #1 (all area treated by LSC) is flat and continuously uniformly distributed on the rail surface, while the hardening layers of the other specimens of #2, #3, and #4 (selectively treated by LSC) are crescent-shaped and isolated.
Figure 5. The cross-section morphologies of the hardening layer.
Figure 6 shows the cross-section microstructure of specimen #2 treated by LSC. The structure was characterized of hardening layer, heat affected zone (HAZ), and substrate, as shown in Fig. 6(a). In Fig. 6(b), small acicular martensite is found in the hardening layer and was produced by rapid heating and cooling of laser quenching, while the growth of austenite was suppressed. The HAZ is composed of martensite, residual austenite, and partial substrate，as shown in Fig. 6(c). The substrate of the rail is composed of typical lamellar pearlite, as shown in Fig. 6(d). Similar structures can be found in specimen #3 and #4.
7
Figure 6. (a) Structure of specimen #2 from surface to substrate; SEM of (b) the hardening layer, (c) the HAZ and (d) the substrate.
After experiment, all the specimens were cut at the middle of the hardening spots, and microhardness distribution was tested along their depth direction, as shown in Fig. 7. The hardness of specimen #1~#4 by LSC is in the range of HV750~850. The hardness of the HAZ is HV400~HV500. The average hardness of the untreated area of the substrate is about HV300.
Figure 7. Microhardness distribution along depth direction of the specimens.
3.2 Wear
8
The weight loss was estimated by weighing specimens before and after the tests. The wear losses of the specimens were measured by an electronic balance (Mettler-Toledo AL204/1) with an accuracy of 0.1 mg. Relative wear resistance R for different specimens is defined by Formula (1) [12]： R=Wb/Wa
(1)
where Wb is the wear losses of the untreated rail specimen after the wear and CFR tests, and Wa is the wear losses of the LSC-treated rail specimen after wear test. Figure 8 shows the wear resistance of the specimens. It is found that the relative wear resistance of specimens treated by LSC is 1.2~40 times higher than that of the untreated specimens, and the relative wear resistance of specimens #1~#4 is 40.3, 9.7, 2.5 and 1.2, respectively, decreasing with the reduction of quenched depth and hardening area ratio.
Figure 8. Wear resistance of the specimens under different processing conditions.
Figure 9(a) shows the surface morphologies of the wear specimens. The widths of the wear patterns on the surfaces of specimens #0 and #4 are much wider than those on specimens #1~#3, as marked by the yellow arrows. This means that the surface plastic deformation of the rail materials is severe. The side wave deformation, pointed out by the blue arrows in specimens #0 and #4, also indicates that both specimens have endured severe deformation. The main reason may be due to that specimen #0 has a low hardness level and specimen #4 has shallow hardening layers and a smaller hardening area ratio.
9
On the contrary, the side wave deformation of specimens #1~#3 with high hardening area ratios is slight or even invisible. It should be noted that not all the LSC-treated specimens have excellent wear resistance. For example, the surface morphology of specimen #4 is similar to the untreated specimen #0. On the other hand, although specimens #2 and #3 have excellent wear resistance, part of the hardening layers under the contact areas are worn away, as shown by the red dashed circles and the residual layers pointed out by black arrows. Some small cracks and spalling can be observed on the wear surface, indicated by the purple arrows. Profilometry was performed to assess the wear depth on the rail specimens (Fig. 9(b)) and to compare and validate the results. On the surface of specimen #0 without laser quenching, an extended groove and deep wear pattern (0.57mm) demonstrate that the wear between the rail and wheel is very serious, which would affect the service life of the rails. In contrast, the wear depth of specimen #1 is very thin. Compared with specimen #0, the wear resistance by LSC treatment (specimens #2~#4) can be improved, but the extent depends on the hardening depths and the area ratios of the quenched dots. The larger the ratios and depths are, the better the wear resistance of the specimens treated by LSC is.
Figure 9. (a) Surface morphologies of specimens after experiment; (b) wear depth of specimens after experiment.
3.3 Contact Fatigue Resistance (CFR) 10
In this research, the depth and length of cracks were analysed to evaluate the CFR of rails. The cross-section morphologies along the longitudinal direction of specimen #0 are shown in Figs. 10(a)-(b), in which there are several cracks with lengths about 10 μm and depths smaller than 50 μm. These shallow and short cracks are prone to be removed by the wear, and no penetrating crack is found in specimen #0. Hence, the service life of the untreated rail is mainly determined by its wear resistance. The cross section of specimen #1 is shown in Figs. 10(c)-(e), in which neither plastic deformation nor a micro-crack can be observed due to its high hardness. However, a crack nucleates and propagates from the hardening layer to the substrate. The leading crack ramifies into several secondary cracks in different directions in the interface between the hardening layer and the substrate, as shown in Fig. 10(c). The depth of the leading crack is 1000 μm, which is 17 times as much as the wear depth (58μm). Further, the downward propagation branch shown in Fig. 10(e) may cause rail damage or fracture during long-term service, which may threaten the transportation safety of the trains. Thus, it should be emphasized that it is the CFR, rather than the wear resistance, that determines the service life of laser all-area hardened rails. This result can explain why whole-surface hardening, either by laser quenching or other traditional quenching techniques, is forbidden in rail strengthening [4-6]
.
11
Figure 10. (a, b) Cracks of untreated specimen #0; (c) crack morphology of specimen #1; (d) microstructure of a leading crack of specimen #1; (e) downward crack propagation of specimen #1; (f) cracks and spalling of specimen #3; (g,h) superficial and horizontal crack of specimen #3; (i,j) cracks of specimen #4.
The cross-section morphology of specimen #3 by LSC is shown in Fig. 10(f), in which the contact fatigue cracks nucleate near the boundary between the hardening layer and the substrate, and then propagate along the surface direction with a length of about 0.1 mm, as shown in Figs. 10(g) and 10(h).
12
Compared with specimen #1, the cracks of specimen #3 are shorter, and their positions are shallower. In Fig. 10(h), the cracks in specimen #3 are about 100 μm and are located on the surface with 38% of the wear depth. Most of the cracks in specimen #3 propagate horizontally, and no penetrating crack is found in the hardening layer or substrate. Similar characteristics are observed in the cross-section morphologies of the specimen #2. Figures 10(i)-(j) display the cross-section morphologies along the longitudinal direction of specimen #4, and it can be observed obviously that specimen #4 has similar plastic deformation and wear resistance to specimen #0. It may be because the hardening spot and depth of specimen #4 are too small, and the surface property is almost the same as the as-received state.
4. Discussion Given the amount of data collected in experiment and in field (section 5), it is necessary to declare in this section that our approach is a preliminary and qualitative interpretation of the results to be confirmed by further detailed analyses, instead of carrying out complex calculations at this stage. Generally speaking, contact fatigue cracks in the rail are caused by surface distortion and relaxation under alternating stresses between the wheel and rail. The fatigue failure is divided into two stages: firstly, the crack nucleates and propagates in the rail
[11-13]
(the simplified two-dimensional plane stress model, Hertz Contact, can
explain the cracking behaviour in the travelling direction
[10]
); secondly, the normal
stress PZ(x) and the tangential stress q(x) from friction are applied on the rail. Given the friction coefficient and the normal stress, q(x) is calculated according to Formula (2) and (3) [14]: Pz(x)=PZMAX[1-(x/b)2]1/2 q(x)=μ·PZ(x)
(2) (3)
where PZMAX is the maximum normal stress, b is the semi axis of the contact area for the wheel and rail, and μ is the friction coefficient between the wheel and rail. The 13
induced stress distribution against the normalized distance from the surface is plotted in Fig. 11(a).
Figure 11. Stress analysis of the rail due to the interaction between the rail and wheel: (a) stress distribution on rail; (b) change of maximum shear stress with different friction coefficient μ [14].
Assuming that both the rail and the wheel have perfectly smooth surfaces and no friction occurs between them, only the normal stress from the wheel exists on the rail. Thus, the maximum shear stress (τmax) on the rail surface can be defined by Formula (4) [11]: τmax={[1/2(σx-σz)]2+τxz}1/2
(4)
According to Formula (4), the shear stress isoline can be drawn, and τmax is found in the sub-surface
[11]
. Here, plastic deformation is prone to take place, and contact
fatigue cracks nucleate. This means that the internal shear stress exists in the rail even on an ideal smooth surface, and the crack nucleation is not in the upper surface but in the sub-surface of the rail where τmax is. In realistic engineering conditions, there is no pure rolling for the wheel and rail 14
pair, and the tangential stress caused by wheel/rail friction cannot be ignored. Multiple studies indicate that τmax increases and depth decreases with an increase of the friction coefficient μ
[14-16]
. Figure 11(b) presents the shear stress distributions
under different friction values μ by computational simulation
[14]
. When the friction
coefficient μ changes from 0 to 0.4, τmax increases from 380 MPa to 450 MPa; the position of the cracks will move from sub-surface (i.e. 1 normalized depth) to the upper surface of the rail [14]. Furthermore, the rails are subjected to the mixed-mode cyclic loading, including mode I stress (caused by the normal stress) and mode II stress (caused by the tangential stress) in real engineering conditions. Since the stress mode II tends to change the original crack propagation directions, the wear and the rolling contact fatigue process will become more complicated
[17]
. The stress intensity factors (SIFs)
for mode I and mode II are combined respectively for an effective stress intensity factor to describe crack propagation rates. Tanaka
[16]
and Weertman
[18]
express a
crack propagation rate, similar to Paris’ Law [11], by Formula (5): da/dN=C(ΔKeff)4, ΔKeff=KI+8KII
(5)
where C is the material constant, ΔKeff is the effective SIF, KI and KII are the SIFs of mode I and mode II, respectively, and da/dN represents the crack length increment per loading cycle. Based on Formula (4) and (5), we can discuss the mechanisms of the main physical factors on contact fatigue resistance and the service life of rails. Apart from the material strength of rails, the nucleation and propagation rates of the contact fatigue cracks are determined by the following two conditions: (1) The magnitude and the depth of τmax. The magnitude of τmax determines the crack nucleation time, and the depth of τmax determines the crack nucleation position. The larger the μ or PZ(x) is, the shallower the crack nucleation position would be, and the quicker the crack nucleation would start up.
15
(2) The magnitude of ΔKeff. The value of ΔKeff determines the propagation rates of the contact fatigue cracks after nucleating, and the crack propagating rate is in proportion with the value of ΔKeff. Now we can explain why the laser all-area hardened rails have low fatigue resistance. In typical rail service conditions, as shown in Fig. 12(a), laser all-area quenching will reduce the friction coefficient
[17]
, and make the site of the maximum
shear stress τmax1 deeper, i.e., the crack nucleation location moves inward from the rail’s upper surface
[19]
. Therefore, once the cracks nucleate from the sub-surface, the
cracks cannot be removed by wear and are prone to propagate into long cracks under alternate loadings, greatly shortening the contact fatigue life as well as the service life of the rails. This can explain why the martensite structure of rails is forbidden worldwide. However, laser surface compositing (LSC) has formed a composite surface with the martensite hardening arrays embedded in the untreated pearlite structure matrix. During service, the ductile pearlite is worn away prior by the wheel, and the hardened martensite dot arrays gradually become prominent on the surface, withstanding the main load from the wheel (shown in Fig. 12(b)). The reduction of the contact areas causes the normal and tangential stress to increase greatly. Under this condition, the τmax at the hardening edge (noted as τmax1) gets bigger, and its position (D2) moves upward, meaning the fatigue cracks would probably nucleate at the boundary between the martensite and the matrix on the rail surface, as shown in Fig. 12(c). When the wheels roll at the middle part of the hardening areas, the contact area increases, and the normal stress τmax2 decreases (τmax2＜τmax1). Since the values of KI2 and KII2 in the middle are smaller than those in the boundary areas, the effective SIF at the middle is also smaller than that at the boundary (ΔKeff2＜ΔKeff1). According to Formula (6), (da/dN)2＜(da/dN)1, i.e., the crack propagation rates at the boundary is much faster than those at the middle. As the nucleating cracks at boundaries are superficial, they are prone to be removed by the wear or spalling effect due to the alternating stress between the wheel/rail. 16
Figure 12. Interactions of wheel/rail under different treatment: (a) crack propagation sketch for laser all-area hardened rail; (b) the early stage of experiment of LSC rail; (c) crack propagation sketch for LSC rail.
Summing up the above analysis, the crack propagation mechanisms of the rails treated by LSC are totally different from that by laser all-area quenched rails, and the martensite structure is no longer the dangerous phases for the rails treated by LSC. The rail service life of LSC is dependent upon the CFR and wear resistance, which can be adjusted and controlled by the area proportion and the depths of the hardening arrays. Generally speaking, the CFR and wear resistance are two interactional and competing factors that determine the service life of rails simultaneously
[20]
. On one
hand, the specimens without laser quenching have good toughness and CFR, but its wear resistance is weak due to its low hardness. On the other hand, laser all-area quenched rails have the highest wear resistance. However, the contact fatigue cracks of the samples nucleate easily in the martensite areas and propagate in big angles 17
(nearly 90°), which is harmful and disrupts the rails. When LSC is utilized, the hardening area ratios and depths change greatly with different processing parameters, which may observably affect the wear resistance and contact fatigue resistance. The specimens with bigger hardening area ratios and deeper hardening layers have better wear resistance but worse fatigue resistance. By designing the hardening area sizes and proportions rationally, the service life of the rails can be enhanced by 2.2-3.5 times without obviously deteriorating their CFR.
5. Application Verification More than 600 groups of rails treated by LSC have been applied in Wuhan Railway Bureau of China, as shown in Fig. 13. Table 4 shows the service life of the railway switches by LSC with different laser processing parameters. The results demonstrate that the service life can be enhanced by 2.2-3.5 times as much as that without laser treatment. The service life of those treated by LSC depends on not only the laser processing parameters, but also the service conditions of the rails. The maximum service life can even be enhanced by 10 times the untreated rails. The detailed results will be published in our next paper [21]. Table 4. The service life by laser surface compositing under different processing parameters. Switch Number
#1
#2
#3
Treatment statement
No treatment
LSC
LSC
Service life enhancement
1.0
2.2
3.5
*The parameters by LSC for railway switch #2 and #3 are corresponding to specimen #2 and #3.
18
Figure 13. The railway switches by laser surface compositing are setup in Wuhan Railway Bureau. (a) The workers are setting up the rail switch. (b) The rail switches after setup.
6. Summary (1) A novel laser surface compositing (LSC) is firstly introduced on the rail surface by selective laser quenching, which produced martensite arrays with a hardness of HV800 on the toughness pearlite rail matrix. (2) Due to the synthetic effect of the high hardness of the martensite arrays and good toughness of the rail matrix, LSC can enhance the service life of the rails by 2.2-3.5 times without significant loss in its contact fatigue resistance, and it has broken through the traditional worldwide restriction that forbids the martensite structure on rail surfaces in the past hundred years.
Acknowledgements This research was financially supported by the national key research and development program 2016YFB1102600, P. R. China. The application verification of this research work was supported by the institute of railway technique of Wuhan Railway Ltd., and the authors are grateful to Prof. Yongfeng Lu for valuable discussions.
References [1] W. Zhong, J.J. Hu, P. Shen, C.Y. Wang, Q.Y. Lius, Experimental investigation between rolling contact fatigue and wear of high-speed and heavy-haul railway and selection of rail material, Wear, 271(2011) 2485-2493. [2] G. Donzella, M. Faccoli, A. Ghidini, A. Mazzù, R. Roberti, The competitive role of wear and RCF in a rail steel, Eng. Fract. Mech., 72(2005) 287-308. [3] D.F. CANNON, K.O. EDEL, S.L. GRASSIE, K. SAWLEY, Rail defects: an overview, Fatigue Fract. Eng. M., 26(2003) 865-886. [4] E. Standard, Railway Applications-Track-Rail, Part 1: Vignole railway rails 46kg/m and above, in: 13674-1 13674-1, 2002. [5] AREMA, Manual for railway Engineering SECTION 2 SPECIFICATION FOR 19
STEEL RAILS, 2010. [6] C. Standard, Technical Specifications for the procurement of 43 kg/m ~ 75 kg/m rails, 2004. [7] K. SAWLEY, K. J, Development of bainitic rail steels with potential resistance to rolling contact fatigue, Fatigue Fract. Eng. M., 26(2003) 1019-1029. [8] Y. Totik, R. Sadeler, H. Altun, M. Gavgali, The effects of induction hardening on wear properties of AISI 4140 steel in dry sliding conditions, Materials and Design, 24(2003) 25-30. [9] R. Reiff, In-Track Demonstration of Laser-Treated Rail to Reduce Friction and Wear, U.S. Department of Transportation, 2007. [10] B. H. Hertz, On the contact of elastic solids,Zeitschrift fur Reine und AngewandteMathematik, 92(1882) 156--171. [11] L.M. Keer, Contact Mechanics (K. L. Johnson), The Society for Industrial and Applied Mathematics, 1987. [12] Q.H.X.Z. Dengzhi Wang, Microstructures and performances of Cr13Ni5Si2 based composite coatings deposited by laser cladding and laser-induction hybrid cladding, J. Alloy. Compd, (2014) 502-508. [13] J.H. Beynon, J.E. Garnham, K.J. Sawley, Rolling contact fatigue of three pearlitic rail steels, Wear, 192(1996) 94-111. [14] J. Seo, S. Kwon, H. Jun, D. Lee, Fatigue crack growth behavior of surface crack in rails, Procedia Engineering, 2(2010) 865-872. [15] B. Trollé, M.C. Baietto, A. Gravouil, S.H. Mai, B. Prabel, 2D fatigue crack propagation in rails taking into account actual plastic stresses, Eng. Fract. Mech., 123(2014) 163-181. [16] K. Tanaka, Y. Nakai, M. Yamashita, Fatigue growth threshold of small cracks, Int. J. Fracture, 17(1981) 519-533. [17] R.J. DiMelfi, P.G. Sanders, Mitigation of subsurface crack propagation in railroad rails by laser surface modification1, Surface and Coatings Technology, 106(1998) 30-43. 20
[18] J. Weertman, Theory of fatigue crack growth based on a BCS crack theory with work hardening, Int. J. Fracture, 9(1973) 125-131. [19] T. Kuminek, K. AnioEk, J. M Yńczak, A numerical analysis of the contact stress distribution and physical modelling of abrasive wear in the tram wheel-frog system, Wear, 328–329(2015) 177-185. [20] D.T. Eadie, D. Elvidge, K. Oldknow, R. Stock, P. Pointner, J. Kalousek, P. Klauser, The effects of top of rail friction modifier on wear and rolling contact fatigue: Full-scale rail–wheel test rig evaluation, analysis and modelling, Wear, 265(2008) 1222-1230. [21] Q.H.Y.Z. Xiaoyan Zeng, Service life under different laser processing parameters, to be submitted to Surface and Coatings Technology.
Highlights
In this study, a novel laser surface compositing (LSC) is firstly introduced on the rail surface by selective laser quenching, which produced martensite arrays with a hardness of HV800 on the toughness pearlite rail matrix.
By designing the hardening area sizes and proportions rationally, the wear resistance of the rails is enhanced by 2.2-3.5 times without noticeably deteriorating their contact fatigue resistance (CFR).
Due to the synthetic effect of the high hardness of the martensite arrays and good toughness of the rail matrix, LSC can enhance the service life of the rails by 2.2-3.5 times, and it has broken through the traditional worldwide restriction that forbids the martensite structure on rail surfaces in the past hundred years.
The mechanisms of LSC to enhance the service life of rails are analyzed systemically, and the rails treated by LSC are taken into application to verify the theoretical research. 21
22