Surface & Coatings Technology 277 (2015) 181–187

Contents lists available at ScienceDirect

Surface & Coatings Technology

j ourna l homepage: www.e lsev ie r .com/ locate /sur fcoat

Effect of one-step laser processed biomimetic coupling units' degrees onrolling contact fatigue wear resistance of train track alloy steel

Wanshi Yang a, Ti Zhou b, Wang Zhang a, Jie Li a, Zhikai Chen a, Fang Chang c, Haifeng Zhang d, Hong Zhou a,⁎a The Key Lab of Automobile Materials, The Ministry of Education, Jilin University, Changchun, Jilin 130025, PR Chinab The School of Mechanical Science and Engineering, Jilin University, Changchun, Jilin 130025, PR Chinac The College of Materials Science and Engineering, Jilin University, Changchun, Jilin 130025, PR Chinad The Department of Mechanical and Automotive Engineering, Changchun University, Changchun 130028, PR China

⁎ Corresponding author.E-mail address: walkermars@163.com (H. Zhou).

http://dx.doi.org/10.1016/j.surfcoat.2015.07.0570257-8972/© 2015 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 25 November 2014Revised 13 April 2015Accepted in revised form 26 July 2015Available online 29 July 2015

Keywords:BiomimeticLaser remeltingRolling contact fatigue wear

The working life of train track is restricted by rolling contact fatigue wear. The rolling contact fatigue wearresistance of train track alloy steel is a majority in ensuring the safety of rail transit and prolonging the life oftrack. To improve the rolling contact fatigue wear resistance of train track alloy steel. The wearable surfacesassembled with different degrees' striation biomimetic coupling units were prepared on the alloy steel by one-step laser remelting processing. The microstructures and micromorphologies of bionic units were characterizedby optical microscope, scanning electron microscopy and field emission scanning electron microscopy. Thespecimens with units were tested for rolling contact fatigue wear resistance. The surface appearances of thetest specimens were compared to each other. By finite element analysis, mechanism of the unit on enhancingwear resistance was studied. A discussion on the wearable surfaces assembled with the different degrees'striation biomimetic coupling units was had. Due to shift of surface stress concentration, the specimens withthe units presented better rolling contact fatigue wear resistance than the blank sample. And it attributed todifferent effective length support ratios that the specimens with striation units of 30° and 60° get better wearresistance than the one with striation units of 45°.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

With the development of rail transit, train is playing a more impor-tant role in traffic. Safety and reliability are two key issues in the railwayfield [1]. The increase of transport capacity will make a higher require-ment on the working life of train track. The improvement on wearresistance of railway surface in a simple and economical way hasbecome a nonnegligible task.

When a train is running on the track, the contact type between therollers of train and railway belongs to line contact. The wear is mainlycaused by rolling contact fatigue (RCF) under the condition. The damagecan lead to spalling of the railhead or complete failure of the rail [2].A previous study from Josefson et al. said the surface-initiated cracksappearing arise due to repeated plastic deformation and consecutiveaccumulated damage on the rail surface [3]. Rolling contact fatigue canbe defined as cracking and pitting/delamination limited to the nearcontact surface of bodies rolling against each other [4]. In the contactcondition, typical damage includes pitting, spalling, cracking and‘squat-like’ damage. There into, pitting is formed in all rolling pairs,and the others belong to the development of pits [5].

A series of research have been carried on to enhance the RCFwear resistance of materials surface. By finite element simulations, theprinciple of fatigue crack initiation under different rolling contactwear condition was discussed [3]. By comparing the cracks growth oftwo kinds of substrate materials on rolling contact fatigue, Franklinet al. studied the mechanism of the wear behavior [6]. At the sametime, the ways of improving the matrix surface wear resistance bylaser processing were proposed. Via unitizing the laser cladding todeposit Ni based amorphous matrix coatings on mild steel substrate,themicrostructure andwear properties ofmaterialswould be improvedsignificantly [7]. Laser surface alloying with silicon was employed tostrengthen the wear resistance of mild steel surface [8].

In nature, creature with the proper shapes and structures can sur-vive andmultiply because of survival of the fittest. The best harmoniousand adaptive system according to optimize and couple the factors con-sist of the best shape, structure and materials [9–11]. The phenomenonis called biomimetic coupling. Based on the principle,many studies havebeen carried out on improving the properties of material. By imitate thebionic morphology of lotus leaves, the super hydrophobic surface wasprepared on aluminum alloy [12]. And the wear resistance can also bereformed by constructing non-smooth surface like shell [13]. Not onlythe mechanical properties, a new sound absorber was developed bythe coupling absorption structure of a typical silent flying bird-owl[14], and the butterfly wings has been well understood as a model

Fig. 1. The SEM of the matrix.

182 W. Yang et al. / Surface & Coatings Technology 277 (2015) 181–187

with outstanding light trapping property [15]. In the past few years,Zhou et al. have studied the fatigue wear resistance of cast iron andtool steel. According to previous reports, the facts that the wearablesurface with units processed by laser can enhance the wear resistanceof tool steel and cast iron under sliding wear condition significantlyhas been proved [16–18]. The surfacewith units presented the alternat-ed hardness as the shape of ground beetles' elytrum or tree leaf, inwhich the units are identified as biomimetic units. The units processedby laser remelting has compacted structure and grain refinement,which results in excellent performance on the sliding wear resistance.Whether the train tracks alloy steel cured by the laser remelting canalso present efficient enhancement on rolling contact fatigue wearresistance is to be expected.

In the paper, biomimetic units were prepared on the specimens oftrain track alloy steel by one-step laser processing. The microstructuresof units were characterized by optical microscope (OM), scanningelectron microscopy (SEM) and field emission scanning electronmicroscopy (FESEM). The stress analysis of the unit under rollingcontact was shown by finite element analysis software. The wearablesurfaces with different degrees' striation units were constructed onthe specimens. Rolling contact fatigue wear test was carried on bya self-building wear tester. The wear resistance mechanism of thewearable surfaces was discussed. The effects of microstructure anddegree of the units on the wear behavior of the surface under rollingcontact were also studied.

2. Experimental

2.1. Materials

The chemical compositions of the train track alloy steel materials aregiven in Table 1. The substrate materials belong to high manganesesteel, and the tensile properties and the microhardness of materialsare also listed in the table.

The Fig. 1 shows the microstructure of the material matrix, which iscomposed of lamellar pearlite.

2.2. Sample preparation

Before laser remelting processing, the experimental samples forrolling contact fatigue wear test were cut from the train track abovewith dimensions of 120 mm length, 15 mmwidth and 5 mm thicknessby a wire electrical discharge machining (Huadong Group, DK77,China), whichwas equippedwith a computer numerical control system.

A solid state Nd: YAG laser of 1064 μm and maximum 300 W wasemployed to fabricate the bionic unit. The temperature in samplesprocessing was 22 °C. The laser head was mounted vertically in theZ-direction and was adjustable. Movement along X and Y axes wasused to process the bionic unit with varied surface shapes while thatalong Z axis was to adjust the desired defocusing amount [19].

The processing parameters were laser duration 5.0 ms, laser inputenergy density 210.39 J/cm2, frequency 5 Hz, defocusing amount14 mm, scanning speed 0.5 mm/s and a circular spot size 3.8 mm indiameter on the specimen surface.

The striation bionic units were assembled on the samples by laserremelting. According to the different degrees between the striationunits and the long edge of the sample, Sample 1 (S1) was fabricatedby the striation units of 30°, while Sample 2 (S2) assembled by thestripes of 45°. The striation units of 60° were spread on Sample 3 (S3).Through controlling the displacement of the workbench, the horizontal

Table 1The chemical compositions (wt.%), the microhardness (HV0.2) and the tensile properties (Mpa

C Si Mn S P Ni

0.77232 0.24486 1.23508 0.02843 0.01435 0.00952

distance of the parallel striation units on the surfaces of the testspecimens was kept on 6 mm, as shown in Fig. 2.

2.3. Characterization method

The cross-section of the unit was obtained after laser remelting pro-cessing, whose microstructure and dimensions were studied by opticalmicroscope (Zeiss, Axio. Imager. A2m, Germany). Themicrohardness inthe cross-section was measured by a Vickers hardness tester (Buehler,5104, USA) with a 0.2 kgf applied load. The component phases ofeach unit morphology and matrix were characterized by scanningelectron microscopy (JEOL, JSM-5600, Japan), field emission scanningelectron microscopy (Zeiss, Supra 40, Germany) and X-ray diffractioninstrument (Fangyuan, DX-2700, China).

2.4. Wear tests

A self-building rolling contact fatigue wear tester was used to servethewear tests, whose schematic diagram is shown in Fig. 3 [20]. The ap-plied loading in the tests was 8 kg. In terms of control variables method,levels of pitting were served as the standard to the RCF wear resistanceof specimens by maintaining a constant time. The rotational speed ofthe electrical machine was 690 rpm, while the RCF wear time was60 h. The wear test condition was built to simulate real train operatingsituation. By comparing difference in mass of the specimen before andafter the experiment, the mass losses of the wear test samples weregiven. Above date was measured by a sensitive electronic balance withan accuracy of 0.0001 g.

3. Result and discussion

3.1. Microstructure

As shown in Fig. 4, X-ray diffraction is used for phase analysis of theunit and matrix. The phase compositions of unit present a mixed phaseof martensite containing carbide and a small amount of ferrite, whilethat of thematrix appear to have the original phase of ferrite. The figurealso shows X-ray full width at half maximum (FWHM). The FWHM ofX-raydiffraction peaks of untreatedmatrix is 0.313°,while that of bionicunits is 0.431°.

) of the materials of train track alloy steel (wt.%).

Fe Hardness Tensile properties Fracture strength

Balance 312.2 968.6 888.25

Fig. 2. The distribution of biomimetic units on the surface of specimen.

Fig. 3. The schematic diagram of the self-building RCF wear tester [16].

183W. Yang et al. / Surface & Coatings Technology 277 (2015) 181–187

According to Scherrer formula:

D ¼ K λβ cos θ

:

Where D is the grain size, K is the Scherrer constant, λ is the wavelength of X-ray, β (rad) is the FWHM of diffraction peak and θ (deg.)is the peak position of diffraction line, and the X-ray diffraction peaksbroadening means grain size or sub grain size refinement. Owing tothe refinement of grains, the better comprehensive properties ofmaterial are obtained [21,22].

In addition, the X-ray diffraction peaks of bionic units exhibited amicrostructure comprising carbide. The appearance of carbide impliesthat the surface with bionic units get an improvement on fatigueresistance in rolling contact condition [23].

In Fig. 5, the cross-sectional view of the bionic unit morphologyprocessed by laser is depicted. Due to the corrosion of nital, the zoneswith different structures demonstrate various shades under opticalmicroscope. The highlight means the zone has a higher resistance tocorrosion and a stronger structure. The cross-sectional shape of unitpresents arc. The dimensions of unit are about 1.87 mm wide and0.44 mm deep. There is a highlight annular section with approximately1 mm wide in the outer of the unit.

Fig. 4. X-ray diffraction line profiles of the biomimetic unit and the matrix.

Fig. 6(a) and (b) show the cross-sectional SEM micrographs ofthe external and internal of bionic unit, respectively. Fig. 6(c) presentsthe FESEM micrograph of junction of external and internal. As seenin the figures, the internal of bionic unit presents lath martensite.While the metallic structure of that in the external of bionic unitattaches to the martensite in grain refinement. In the area of externalbionic unit, some granulated carbides spreads. Carbides could attributeto the fairly rapid solidification, which prevents the reinforced particlesdissolve into the matrix. And carbides also contribute to limit thegrowth of cracks during the wear process [24].

The different areas possess difference onmetallurgical behavior. Thewhole arc zone is called remelted zone, while the rest is matrix. Afterthe laser beam incidence at a high heating rate and the quench coolinginfluence of surrounding, the remelted zone formed through a processof solidify and recrystallization. On account of thermal conductivity ofthe material and difference of temperature between the matrix andouter air environment, two kinds of structures, different in grain sizes,were developed in the external and internal of bionic unit.

3.2. Microhardness

Fig. 7 shows the microhardness of bionic unit. Before a rapid in-creasing to the peak of 950 HV approximately, the value of hardnesspresents a gentle fluctuation. After a significant rise, the coefficientbegins to flatten again from the depth of ~400 μm. Here, the valuesof the first flat area and the peak area are both higher than the secondflat area. The two zones with increasing hardness of the biomimeticunit could be attributed to the formation of fine martensite grains,the precipitation of carbides and the increase of dislocation density[19]. Due to the differences in the grain size and structure betweenthe two zones, the hardness presents the divergence. The situationof microhardness in the transformation zone is not showed fromthe figure, but it makes the transition to that of substrate directly.What above suggests is that the addition of the biomimetic unitshas a positive effect on the increasing of surface microhardness,with a potential for wear resistance.

3.3. Results of the RCF wear tests

The average mass losses of the specimens with bionic units as afunction of the degrees between the striation units and the long edgeof the sample are presented in Fig. 8. It can be shown that the wear

Fig. 5. The cross-sectional view of the bionic unit morphology.

Fig. 7. The average microhardness on the cross section of the unit.

184 W. Yang et al. / Surface & Coatings Technology 277 (2015) 181–187

losses of biomimetic specimens, compared with the untreated samplein the same condition of RCF wear, decrease significantly. Wear lossesof S1 reduce about 85% among them, while that of S3 declines by86.3%. It is close to S1. The specimen assembled with 45° striationunits (S2) cuts down the mass loss approximate 73.9%. All these showthat the biomimetic specimens with the striation units possess betterrolling contact fatigue wear resistance than untreated material.

Themorphologies of theworn surfaces of the biomimetic specimenswith different degrees' units have been compared in Fig. 9 underthe same RCF wear condition. It is shown that the pitting wear isthe dominative wear mechanism of the specimens. As demonstratedin the figure, the biomimetic specimens present relative slighterwear, compared with the untreated one appearing serious fatiguepitting and wear. There into, better performances of wear resistanceamong the specimens are shown on S1 and S3. Above is provedthat biomimetic units can resist the removal of materials andretard [18].

Fig. 6. The cross-sectional SEM micrographs of the external (a) and internal (b) of bionic

3.4. Effects of biomimetic units

To explain themechanism on RCF wear resistance of the bionic unit,the stress analysis' models with sole unit pressed by a single roll and theblank model without the unit were constructed (as shown in the insertof Fig. 11). Fig. 10 demonstrates the stress moiré diagrams of themodelwith bionic unit ((b) and (d)) and the untreated one ((a) and (c)). Theapproximate global size of meshing is 0.7. The bright level of meshstands for the level of stress concentration. The boundary of stress onthe surface is marked by white lines. Previous works have pointed outthat fatigue failures in service invariably occur at stress concentrations[25]. Of which the red region is the extreme value of stress, and it alsoindicates the fatigue pitting or wear occurs easily near the region.

unit, the cross-sectional FESEMmicrographs of junction of external and internal (c).

Fig. 8.Wear mass loss as a function of units' angle.

185W. Yang et al. / Surface & Coatings Technology 277 (2015) 181–187

By comparison of the ranges of white mark, the addition of bionicunit reduces the area and the level of stress concentration rangeon the surface of the specimen obviously, which is shown inFig. 10(a) and (b). The cross-sectional views of the diagrams arepresented in Fig. 10(c) and (d).

To discuss the influence of units on stress concentration, a series ofpoints were generated along the direction of the arrows as the insertof Fig. 11 shown. The relation curves between sectional stress andpoints' depth of two types of models are shown in the figure. Comparedwith the substrate, there is amore refinedmicrostructure in the units. Itleads to bettermechanical properties, which results in thematerials canbear greater stress. Based on previous discussion, the unit presents abetter potential on RCF wear resistance because of the microhardnessand microstructure. Therefore, the units can support heavier load thanuntreated matrix. Fig. 11 also shows that the stress value near the sur-face appears a significant drop on account of the addition of bionicunits, with a reducing of the range of the surface stress. The depth of

Fig. 9. The morphologies of the worn surfaces: (a)

stress concentration shifts to the subsurface, and the concentrationnear the surface could be relieved. Above all, the bionic structure playsan important role on limiting the concentration region to spread. Itachieves the region to the deep and makes it confined in the area ofthe unit. The result shows that the bionic structure has a significantimprovement on stress concentration reducing of the area.

3.5. Effects of striation units' degrees

With a view to further discuss the effect of striation units' degrees onRCF resistance, the sketch map of the RCF test behavior on biomimeticspecimen comes into Fig. 12. The contact type between rollers andthe specimen belongs to line contact. When the rollers runs on thespecimen along the arrow direction, the bionic units provide supporton the contact line. The support can enhance the RCF resistance of thespecimen. The ratio of the effective supporting length of units reflectingon the contact line is defined as the effective length support ratio (Re).The degree between the striation units and the long edge of the sampleis θs. The lengthof a roller is L0. Lwhas been set as the overallwidth of theunits, so the effective length (Le) of striation units on the rollers isLwtan θs.

According to the parameters above and the definition of the effectivelength support ratio, Re is present as follow:

Re ¼ Lwtan θsL0

� 100% :

Re is an interval value, which is illustrated in Table 2. It can beseen that the range of Re30 equals to that of Re60, and the upperand lower bound of them are both larger than the bounds of Re45.It coincides with the result of wear tests that S1 and S3 possess thelower wear losses than S2, which implies the higher ratio means abetter RCF resistance.

untreated specimen, (b) S1, (c) S2, and (d) S3.

Fig. 10. The stress moiré diagrams of model with the unit (b, d) and without the unit (a, c).

186 W. Yang et al. / Surface & Coatings Technology 277 (2015) 181–187

4. Conclusions

Using one-step laser remelting processing, a kind of biomimeticunits was prepared on the surface of alloy steel from train tracks. Themicrostructure of the units was investigated, and the stress analysis ofthat under rolling contact condition was discussed by finite elementanalysis method. The different degrees' striation units were assembledon the test specimens, and then wear test was done. On the basis ofthe theory of effective length support ratio (Re), the influence of units'degree on rolling contact fatigue wear resistance was studied.

1) The SEM and the FESEM images indicated that the substrate micro-structure of materials belong to pearlite, which manufacture most

Fig. 11. The sectional stress curvesof theunit andblank. Insert offigure: the stress analysis'models of sole units pressed by a single roller.

rail steel and is significantly altered by rolling-sliding contact [26].There is inter-outer structure in the units, and lath martensite playsamajor role in themicrostructure. Ofwhich, the outer gets a better re-finement and some granulated carbides appeared in it. Comparedwith the matrix, a significant improvement on the microhardness ispresented in theunits. The surface treatment (remelting and cladding)on stress distribution and damage simulation was studied in progress[27]. Above indicators demonstrate a positive effect on fatigue wear.

Fig. 12. The sketch map of the RCF test behavior on the specimen with the striation units.

Table 2The effective length support ratios of the specimens with differentdegrees' striation units.

Degree(°)

Effective length support ratio(Re)

30 29.92%–35.44%45 25.45%–34.55%60 29.92%–35.44%

187W. Yang et al. / Surface & Coatings Technology 277 (2015) 181–187

2) From the finite element stress analysis, it contributes to relieve thestress concentration that the depth of stress concentration shifts tothe subsurface due to the addition of units.

3) According to the wear test result, the samples with different degrees'striation units are better than the untreated one on fatiguewear resis-tance. There into, specimens with 30° units and 60° units process amore significant improvement than that of 45° units, mainly becausethe hard lengths of 30° units and 60° units reflecting on the contactline are greater than that of 45° so that larger effective length supportratios are illustrated. The enhancement of striation units in differentdegrees on fatigue wear resistance has been accepted [28].

Acknowledgments

This article was supported by Project 985 — High PerformanceMaterials of Jilin University and the Project 985 — Bionic EngineeringScience and Technology Innovation and National Natural ScienceFoundation of China (No. 51275200).

References

[1] A. Cavuto, M. Martarelli, G. Pandarese, G.M. Revel, E.P. Tomasini, Experimentalinvestigation by laser ultrasonics for high speed train axle diagnostics, Ultrasonics55 (2015) 48–57.

[2] R.I. Carroll, J.H. Beynon, Decarburisation and rolling contact fatigue of a rail steel,Wear 260 (2006) 523–537.

[3] J.W. Ringsberg, M. Loo-Morrey, B.L. Josefson, A. Kapoor, J.H. Beynon, Prediction offatigue crack initiation for rolling contact fatigue, Int. J. Fatigue 22 (2000) 205–215.

[4] R. Nieminen, P. Vuoristo, K. Niemi, T. Mäntylä, G. Barbezat, Rolling contact fatiguefailure mechanisms in plasma and HVOF sprayed WC-Co coatings, Wear 212(1997) 66–77.

[5] O.P. Datsyshyn, V.V. Panasyuk, Pitting of the rolling bodies contact surface,Wear 251(2001) 1347–1355.

[6] F.J. Franklin, G.J. Weeda, A. Kapoor, E.J.M. Hiensch, Rolling contact fatigue and wearbehaviour of the infrastar two-material rail, Wear 258 (2005) 1048–1054.

[7] R.F. Li, Z.G. Li, J. Huang, Y.Y. Zhu, Microstructure and wear properties of laserprocessed Ni based amorphous matrix coatings, Surf. Eng. 28 (2012) 513–516.

[8] J.D. Majumdar, Development of in-situ composite surface on mild steel by lasersurface alloying with silicon and its remelting, Surf. Coat. Technol. 205 (2010)1820–1825.

[9] B. Ji, H. Gao, A study of fracture mechanisms in biological nano-composites via thevirtual internal bond model, Mater. Sci. Eng. A 366 (2004) 96–103.

[10] A.B. Kesel, U. Philippi, W. Nachtigall, Biomechanical aspects of the insect wing: ananalysis using the finite element method, Comput. Biol. Med. 28 (1998) 423–437.

[11] J. Lomakin, Y. Arakane, K.J. Kramer, R.W. Beeman, M.R. Kanost, S.H. Gehrke,Mechanical properties of elytra from Tribolium castaneum wild-type and bodycolor mutant strains, J. Insect Physiol. 56 (2010) 1901–1906.

[12] Y. Liu, J. Liu, S. Li, Y. Wang, Z. Han, L. Ren, One-step method for fabrication ofbiomimetic superhydrophobic surface on aluminum alloy, Colloids Surf.A Physicochem. Eng. Asp. 466 (2015) 125–131.

[13] L. Tian, X. Tian, Y. Wang, G. Hu, L. Ren, Anti-wear properties of the molluscan shellScapharca subcrenata: influence of surface morphology, structure and organicmaterial on the elementary wear process, Mater. Sci. Eng. C 42 (2014) 7–14.

[14] Y. Wang, C. Zhang, L. Ren, M. Ichchou, M.-A. Galland, O. Bareille, Sound absorption ofa new bionic multi-layer absorber, Compos. Struct. 108 (2014) 400–408.

[15] Z. Han, S. Niu, M. Yang, J. Zhang, W. Yin, L. Ren, An ingenious replica templated fromthe light trapping structure in butterfly wing scales, Nanoscale 5 (2013) 8500–8506.

[16] D.L. Cong, H. Zhou, Z.A. Ren, L.Q. Ren, G.Y. Liu, B.S. Lu, C. Meng, C.W. Wang, Thermalfatigue resistance of H13 die steel repaired by partial laser surface remeltingprocess, Mater. Sci. Technol. 30 (2014) 355–362.

[17] Q.-c. Guo, H. Zhou, C.-t. Wang, W. Zhang, P.-y. Lin, N. Sun, L. Ren, Effect of mediumon friction and wear properties of compacted graphite cast iron processed bybiomimetic coupling laser remelting process, Appl. Surf. Sci. 255 (2009) 6266–6273.

[18] H. Zhou, N. Sun, H. Shan, D. Ma, X. Tong, L. Ren, Bio-inspired wearable characteristicsurface: wear behavior of cast iron with biomimetic units processed by laser, Appl.Surf. Sci. 253 (2007) 9513–9520.

[19] C. Wang, H. Zhou, Z. Zhang, Y. Zhao, D. Cong, C. Meng, P. Zhang, L. Ren, Mechanicalproperty of a low carbon steel with biomimetic units in different shapes, Opt. LaserTechnol. 47 (2013) 114–120.

[20] W. Yang, H. Zhou, L. Sun, C. Wang, Z. Chen, Effect of biomimetic coupling units'morphologies on rolling contact fatigue wear resistance of steel from machinetool rolling tracks, Opt. Laser Technol. 57 (2014) 175–180.

[21] S. Lian, L. Chenglao, Effect of laser melting processing on the microstructure andwear resistance of gray cast iron, Wear 147 (1991) 195–206.

[22] C. Meng, H. Zhou, D. Cong, C. Wang, P. Zhang, Z. Zhang, L. Ren, Effect of biomimeticnon-smooth unit morphology on thermal fatigue behavior of H13 hot-work toolsteel, Opt. Laser Technol. 44 (2012) 850–859.

[23] X.C. Zhang, B.S. Xu, S.T. Tu, F.Z. Xuan, H.D. Wang, Y.X. Wu, Fatigue resistance andfailure mechanisms of plasma-sprayed CrC–NiCr cermet coatings in rolling contact,Int. J. Fatigue 31 (2009) 906–915.

[24] K. Shi, S. Hu, H. Zheng, Microstructure and fatigue properties of plasma transferredarc alloying TiC–W–Cr on gray cast iron, Surf. Coat. Technol. 206 (2011) 1211–1217.

[25] D. Taylor, M. Hughes, D. Allen, Notch fatigue behaviour in cast irons explained usinga fracture mechanics approach, Int. J. Fatigue 18 (1996) 439–445.

[26] J.E. Garnham, C.L. Davis, The role of deformed rail microstructure on rolling contactfatigue initiation, Wear 265 (2008) 1363–1372.

[27] S.M. Zakharov, I.G. Goryacheva, Rolling contact fatigue defects in freight car wheels,Wear 258 (2005) 1142–1147.

[28] Z.-k. Chen, S.-c. Lu, X.-b. Song, H. Zhang, W.-s. Yang, H. Zhou, Effects of bionic unitson the fatigue wear of gray cast iron surface with different shapes and distributions,Opt. Laser Technol. 66 (2015) 166–174.