Friction and wear performance of laser peen textured surfaceunder starved lubrication

Kangmei Li a, Zhenqiang Yao a,n, Yongxiang Hu a, Weibin Gu b

a State Key Laboratory of Mechanical System and Vibration, School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, Chinab Shanghai Aircraft Design and Research Institute, Shanghai 201210, China

a r t i c l e i n f o

Article history:Received 28 January 2014Received in revised form3 April 2014Accepted 12 April 2014Available online 23 April 2014

Keywords:Surface texturingLaser processingFriction testStarved lubrication

a b s t r a c t

Laser peen texturing (LPT), which is a new surface texturing technology, was proposed to fabricatemicro-dimple arrays on copper surface. Pin-on-disk experiments were conducted under different normalloads and sliding speeds to investigate friction and wear behaviors of untextured and textured samplesin starved lubrication. It was found that friction performance of the surface is improved significantlyafter LPT. Microscope observation and EDS analysis showed that textured surface could reduce bothabrasive and adhesive wear as compared to untextured surface. Results also indicated that an optimumtexture density might exist at which surface shows the best friction and wear behavior.

& 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Surface texturing as an approach for controlling the friction andwear behavior of tribological components is well known for manyyears. As early as 1966, the effect of surface texturing on lubrica-tion was studied by Hamilton et al. [1]. Now surface texturing hasbeen widely recognized as a viable means to reduce frictioncoefficient, improve load carrying capacity and wear resistanceof tribological systems. Micro-dimple array is a typical kind oftexture and plays an important part in improving tribologicalperformance. In starved lubrication, dimples act as reservoirs oflubricant, which supply back-up lubricant when lubricant film isbroken down [2]. In hydrodynamic lubrication, dimples areutilized as micro-hydrodynamic bearings to help improve loadcarrying capacity of the lubricant film [3]. Dimples could also serveas traps for wear debris in both lubricated and dry sliding [4].

Many studies on various forms, sizes and shapes of micro-dimples on sliding surfaces for tribological applications have beenreported over the last two decades [2,5–11]. Ryk et al. [2] studiedthe effect of surface texturing under starved lubrication condi-tions, and found that with optimum dimple depth and lowlubricant viscosity the texturing was beneficial over the entirerange of tested flow rates. However, with the deepest dimples or

with high lubricant viscosity the surface texturing may be detri-mental under certain operating conditions. Galda et al. [8] studiedthe effect of surface texturing on lubrication regime transitionsfrom mixed to hydrodynamic and found that the shape anddistribution of oil pockets are the main factors affecting thelubrication kinds. Higuera Garrido [10] studied tribological beha-vior of laser-textured NiCrBSi coatings. Results demonstrated thestrong correlation among texture density, dimple diameter andcontact area for reduction of the friction coefficient. Tang et al. [11]fabricated multi-dimples using a miniature engraving and studiedthe effect of surface texturing on friction and wear under hydro-dynamic lubrication, they concluded that a 5% optimal dimple area

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/triboint

Tribology International

Fig. 1. Schematic of laser peen texturing.

http://dx.doi.org/10.1016/j.triboint.2014.04.0170301-679X/& 2014 Elsevier Ltd. All rights reserved.

n Correspondence to: Room 726, Building A, 800 Dongchuan Road, MinhangDistrict, Shanghai 200240, China. Tel./fax: þ86 21 34206315.

E-mail addresses: lkm718@gmail.com (K. Li), zqyaosjtu@gmail.com (Z. Yao),huyx@sjtu.edu.cn (Y. Hu), guweibin@comac.cc (W. Gu).

Tribology International 77 (2014) 97–105

fraction can generate the greatest hydrodynamic pressure com-pared with other fractions and can reduce friction and wear up to38% and 72%, respectively.

Nowadays, methods to fabricate surface textures could begenerally classified into mechanical [12–16], ion beam texturing[17,18], etching [19–22] and energy beam techniques [23–29].Among these methods, laser surface texturing (LST) is consideredas the most promising texturing technology. The main reason isthat textures fabricated by LST could be precisely controlled andthis process is friendly to the environment [7]. In last few years,many studies have focused on LST, and it has been successfullyapplied on piston rings [24,25], mechanical seals [26,27], hydro-static gas seals [28], journal bearings [29], thrust bearings [30],and soft elasto-hydrodynamic lubrication [23,31,32]. However,laser peen texturing (LPT), which is another energy beam technol-ogy for surface texturing, has been given little attention.

Compared with LST, besides the merits of LST mentionedabove, LPT has more advantages: Firstly, LPT utilizes laser-shock-induced mechanical effect rather than thermal effect, so it couldeffectively avoid the negative influence of ablation in materialsurface integrity. However, during LST, micro-dimples are fabri-cated by a laser ablation mechanism. Tensile residual stress mayoccur in material due to the high temperature during ablation [33].This change in surface integrity will shorten the fatigue life of thematerial [34]; Secondly, after LPT, there is no apparent pile-uparound the micro-dimple which is generally detrimental to thetribological performance, thus post-texturing lapping process forthe purpose of removing pile-up could be omitted. Nevertheless,the lapping process is usually needed for LST according to thestudies of Etsion [8], Yu et al. [27] and Kovalchenko et al. [35].Thirdly, the great shock pressure of LPT can induce deep compres-sive residual stress and hardened layer in material surface andsubsurface, which can improve fatigue life of the material drama-tically [36,37]. The disadvantage of LPT compared with LST may bethat the equipments and the process procedures of LPT are morecomplex.

Table 1Texture parameters used in this study.

Specimen no. 0 1 2 3

Texture density, ρt (%) – 5 13 35Dimple depth, h (μm) – 10 10 10Dimple diameter, d (μm) – 1100 1100 1100

Fig. 2. Schematic of micro-dimple array.

Fig. 3. Surfaces of the specimens.

K. Li et al. / Tribology International 77 (2014) 97–10598

In this study, LPT was adopted to fabricate micro-dimples oncopper surfaces, and the main purpose of this paper is toinvestigate the friction behavior and wear performance of laserpeen textured surfaces under different normal loads and rotationalspeeds in starved lubrication.

2. Experimental details

2.1. Specimen preparation

All textured specimens in this study were processed by LPT.The schematic of LPT is shown in Fig. 1. The surface to be texturedis first coated with an opaque coating (such as black insulationtape) and then covered by a transparent overlay (such as flowingwater). When the laser with high power density (typically GW/cm2) irradiates on the specimen surface, opaque coating melts andvaporizes, and then forms the steam particles. Steam particlescontinue absorbing the laser energy, and then ionize to high-temperature and high-density plasma. Plasma expands rapidly inthe direction away from the specimen and is trapped between thesample and the transparent overlay, creating an extremely highpressure which propagates into the material as a shockwave [38].When the pressure of the shockwave which is on the order of GPaexceeds the dynamic yield strength of the material, a micro-

dimple is fabricated. From the above process, it can be seen thatLPT is a purely mechanical process, which can avoid heat effect ofthe laser.

LPT experiments were performed using a Q-switched Nd:YAGlaser with a wavelength of 532 nm and the Full Width at HalfMaximum (FWHM) of the pulse was about 10 ns. The frequency ofthe laser was 10 Hz. All the textured specimens were processedunder the same LPT parameters: The laser spot diameter was1.3 mm, the repeated shock number for every micro-dimple was1 and the laser power density was 6.76 GW/cm2.

Specimens were prepared from Oxygen-Free High Conductivity(OFHC) copper plate with dimensions of 50 mm�50 mm�2 mm(length�width� thickness). The hardness of the OFHC copperplate was 97 HV (with 25 g load and 20 s hold time) and theaverage roughness of the surface was Ra 0.14 μm. Table 1 lists thetexture parameters used in this research. Specimens from no.1 tono. 3 were textured with different texture densities, while speci-men of no. 0 was an untextured surface used for comparison withtextured ones.

Under the assumption that d is the diameter of the dimple andl is the side length of the micro-dimple unit (see Fig. 2), then thetexture density ρt can be expressed as

ρt ¼SdSt

� 100%¼ πd2

4l2� 100% ð1Þ

where Sd is the area of the dimple, Sd ¼ πd2=4, and St is the area ofthe micro-dimple unit, Sd ¼ l2.

As shown in Fig. 3, the geometric and dimensional character-istics of micro-dimples show good repeatability, indicating thatLPT is an effective method to fabricate controlled micro-dimplearray. The 3D topography and 2D profile of typical micro-dimpleobserved by 3D optical surface profilometer (KS-1100, produced byJapanese Keyence company) are shown in Fig. 4.

2.2. Friction and wear tests

Friction and wear tests were both conducted using a CETR UMT-3tribometer and the pin-on-disk test mode was chosen. Fig. 5(a) and(b) shows the tribometer and the schematic diagram of the contactconfiguration, respectively. The pin was a 10 mm diameter GCr15steel with the hardness of 940 HV (with 25 g load and 20 s hold

Fig. 4. Topography and profile of single micro-dimple.

Fig. 5. Setup used for friction and wear tests. (a) UMT-3 tribometer and (b) schematic diagram of contact configuration.

Table 2Properties of the lubricant used in this study.

Parameters Units Values

ISO viscosity grade – 32Flash point 1C 208Pour point 1C �27Viscosity in 40 1C mm2/s 32Viscosity in 100 1C mm2/s 5.3Density in 40 1C kg/m3 873

K. Li et al. / Tribology International 77 (2014) 97–105 99

time). The stationary pin was pressed at the required load and slidedagainst the rotating disk textured by LPT. The rotational radius of thepin was set as 5 mm. Before the tribilogical test, the sliding surface ofthe pin was polished to mirror smoothness with the average rough-ness of 0.05 μm. After polishing, the pin and the disk were cleaned inacetone solution by ultrasonic cleaning to remove the metal frag-ments and other attachments on metal surfaces.

Lubricant used in all tests was Mobil Vacuoline 1405 lubricant.This lubricant is widely used on guide and chute of machine toolsbecause of its good performance in eliminating stick-slip and chatterduring sliding. The properties of the lubricant are listed in Table 2.

In this research, two sets of friction tests were designed. The firstset was performed with constant rotational speed and linearlyincreased normal load under two different loading rates. The secondset was carried out with constant normal load and step increased

rotational speed. After friction tests, wear tests were conducted tostudy the wear resistance of the specimens. The detailed experi-mental parameters used in this study are summarized in Table 3.Before each friction and wear test, sufficient lubricant was droppedon the contact area, and excessive lubricant was removed from thespecimen by using a rubber blade. After the removal of lubricant,only a thin layer of lubricant appeared on the surface of untexturedspecimen. For textured specimens, due to the existence of micro-dimples, the major proportion of lubricant was stored in the dimplesand a minor proportion remained on the untextured areas. Due tothe difference of texture density, lubricant stored in dimples wasdifferent while the thickness of lubricant above the untextured areawas approximately the same. As a result, it may be considered thatthe lubrication condition for each specimen was approximately thesame except the lubricant volume stored in dimples. In this way, thedifferent friction performance among different specimens couldreflect the effect of texture density. Each friction test was conductedthree times to minimize data scattering.

3. Results and discussion

3.1. Friction behavior under different normal loads

During friction test 1, the normal load increased linearly from20 N to 400 N with the loading rate of 1.27 N/s and 2.54 N/s, whilethe rotational speed kept constant of 150 rpm. For all cases in thistest, the trends of the friction coefficient and normal load curvesare very similar. Typical curves of friction coefficient and normalload are shown in Fig. 6. It is found that the coefficient of frictionkeeps above 0.1 during the whole test. For this reason, and due to

Table 3Parameters of friction and wear tests.

Parameters Units Values

Friction test 1 Friction test 2 Wear test

Rotational speed rpm 150 50–900 (step increased) 150Sliding speed m/s 0.08 0.03–0.47 0.08Normal load N 20–400 (linearly increased) 50 20–280 (linearly increased)Contact pressure MPa 0.26–5.10 0.65 0.26–3.64Loading rate N/s 1.27, 2.54 – 1.3Temperature 1C 2072 2072 2072

Fig. 6. Typical friction coefficient curve under linearly increased normal load.

Fig. 7. Effect of normal load on friction coefficient with loading rate of 1.27 N/s.

Fig. 8. Effect of normal load on friction coefficient with loading rate of 2.54 N/s.

K. Li et al. / Tribology International 77 (2014) 97–105100

the low rotational speed as well as the starved lubrication, it canbe assumed that the sliding contact in this test was boundarylubricated. Moreover, the variation of friction coefficient could bedivided into two stages. During the first stage, the frictioncoefficient seems insensitive to the increase of normal load, thusthis stage can be called stable stage. When normal load reaches acertain value, however, friction coefficient increases rapidly withobvious noise and vibration, indicating that the lubricant filmcollapses and the lubrication fails, therefore, this stage is calledfailure stage. During the friction test, the time at which the frictioncoefficient increases sharply is defined as the failure time, and thecorresponding normal load at the failure time is called themaximum normal load (as marked in Fig. 6) which can reflectthe load carrying capacity of the surface.

The variations of friction coefficient under loading rate of1.27 N/s are shown in Fig. 7. There are two points worth noting:

(1) With the linear increase of normal load, the untextured sur-face failed earlier than the textured ones, and the frictioncoefficient of the untextured surface in stable stage is rela-tively high. The result demonstrates that surface texturinghelps to improve friction performance of boundary-lubricatedsurface. It probably benefits from the so-called secondarylubrication effect. That is, lubricant reserved in the dimplesacts as a secondary source of lubricant and could be drawninto the interface during the relative sliding movement.

(2) Surface with texture density of 13% (abbreviated as surfacetex13%) exhibits the best friction performance as compared tosurfaces with texture densities of 5% and 35% (abbreviated assurface tex5% and tex35%, respectively). In other words, sur-face tex13% has the lowest friction coefficient and the longestfailure time. This result suggests that although the surfacewith micro-dimples is helpful for the improvement of tribo-logical performance compared to untextured, it does not meanthat larger texture density leads to the better tribologicalperformance. There might be an optimum texture density,which leads to the best tribological performance underboundary lubrication. Compared to the surface with appro-priate texture density, the surface with too low or too hightexture density may lead to relatively high friction coefficientand short failure time. Too low texture density means smallnumber of dimples with very limited lubricant and longdistance among dimples. Therefore, the lubricant volumemight be inadequate to cover the whole surface, leading tohigh coefficient of friction and short failure time. On the otherhand, the surface with too high texture density requires great

transverse force for the relative sliding movement and conse-quently causes high friction coefficient. Moreover, too hightexture density indicates limited contact area and great con-tact pressure, therefore leads to large deformation and smallgap between contact surfaces, which makes the lubricant filmthin and easy to collapse.

Fig. 9. Effect of loading rate on friction coefficient.Fig. 10. Effect of loading rate on friction coefficient, failure time and maximumnormal load. (a) Friction coefficient, (b) failure time and (c) maximum normal load.

K. Li et al. / Tribology International 77 (2014) 97–105 101

Fig. 8 shows the effect of normal load with loading rate of2.54 N/s on friction coefficient. It is found that surface tex13% anduntextured surface show the best and worst friction performance,respectively. Furthermore, it could be seen from Fig. 8 that thefriction process of surface tex13% is much more stable than thoseof other surfaces.

In order to compare the effect of loading rate on frictionperformance, the variation of the friction coefficient under twoloading rates are shown in one figure as shown in Fig. 9. It is foundthat the general trend of the friction coefficient variations underloading rate of 2.54 N/s and 1.27 N/s are similar. For a given texturedensity, high loading rate leads to large friction coefficient andshort failure time. In addition, since high loading rate is morelikely to result in an unstable friction process, the oscillation offriction coefficient under loading rate of 2.54 N/s is stronger thanthat under loading rate of 1.27 N/s especially for the surface tex5%,tex35% and the untextured surface.

Quantitative comparisons of friction coefficients, failure timesas well as maximum normal loads for different texture densitiesunder different loading rates are shown in Fig. 10. The datain Fig. 10(a) are obtained by averaging friction coefficients instable stage. It is seen that for surface tex13% with loading rateof 1.27 N/s, the friction coefficient is reduced most significantly bymore than 1/2 compared with untextured surface. Fig. 10(b) showsthat the failure time of surface tex13% is about 3 times longer thanthat of untextured surface under loading rate of 1.27 N/s, andwhen the loading rate increases to 2.54 N/s, this multiple becomesas high as 5 times. Fig. 10(c) presents the maximum normal loadsgained from the fitting curves of raw data. It is found that loadcarrying capacity of the surface tex13% is improved by nearly4 times than that of the untextured one when the loading rate is2.54 N/s. Moreover, it can also be seen from Fig. 10(c) that for acertain texture density, the maximum load carrying capacitywas smaller at bigger loading rate. For this reason, as shown inFig. 10(b), the failure time of each surface is shortened several timeswhen the loading rate increases twice (from 1.27 N/s to 2.54 N/s).

3.2. Friction behavior under different rotational speeds

In friction test 2, the rotational speed was step increased from50 rpm to 900 rpm and the normal load was kept constant at 50 N.Friction coefficient curves of all specimens in this test have similar

trend. So only a typical variation of friction coefficient with stepincreased rotational speed is shown in Fig. 11. At the beginning ofthe test, the friction coefficient is about 0.1 when the rotationalspeed is 50 rpm, as the rotational speed increases to 900 rpm, thefriction coefficient reduced to about 0.05. According to the range ofthe friction coefficient, it could be deduced that the lubricationregime remains boundary and mixed lubrication (that is starvedlubrication), rather than transferring to hydrodynamic lubrication[39]. From the trend of the curve, it is seen that the coefficient offriction becomes smaller and the oscillating of the curve becomesweaker progressively with the increase of rotational speed. Thistrend is in good correlation with the typical Stribeck curve.

For starved lubrication, as shown in Fig. 12(a), the lubricant filmis too thin to separate the two solid surfaces completely, so someopposing asperities touch each other. The friction coefficient nowdepends on both the solid–liquid friction and the solid–solidfriction [39]. By assuming that A represents the total real contact

Fig. 11. Typical friction coefficient curve under step increased rotational speed.

Fig. 12. Schematic of starved lubrication. (a) Under low sliding speed and (b) underhigh sliding speed.

K. Li et al. / Tribology International 77 (2014) 97–105102

area between two sliding surfaces, friction force F could begiven by

F ¼ A½αwτsþð1� αwÞτL� ð2Þ

where αw is the ratio of the solid contact area Am to the total realcontact area A, αw ¼ Am=A. τs is the shear strength of solid surface,and τL is the shear strength of fluid surface. Then the frictioncoefficient of starved lubrication f SL could be obtained

f SL ¼FN¼ A½αwτsþð1�αwÞτL�

N¼ A½αwðτs�τLÞþτL�

Nð3Þ

where N represents the normal load which is a constant in thistest. From Eq. (3), it can be seen that A, τs, τL and N are allconstants here, so there exists proportional relationship betweenfSL and αw.

With the increase of rotational speed, as shown in Fig. 12(b),the lubricant film becomes thick (from h1 to h2) and less asperitiestouch each other, resulting in small area of solid–solid contact andsmall value of αw [39,40]. As a result, the friction coefficient ofstarved lubrication fSL shows a gradual decreasing trend in Fig. 11.

Effect of texture density on the friction coefficient under stepincreased rotational speed and constant normal load is shown inFig. 13. It is seen that for all specimens tested, friction coefficientsdecrease with the increase of rotational speed, but the frictioncoefficient of the untextured surface is higher than that of texturedsurface. Moreover, the smallest friction coefficient is also obtainedby the surface tex13%, followed by the surface tex35% and tex5%.These results are consistent with those under the constant rota-tional speed and linearly increased normal load in friction test 1.Therefore, it is confirmed that texture density of 13% is the bestamong the three texture densities in this study.

3.3. Analysis of worn surfaces

Micrographs in Fig. 14(a)–(d) shows the worn surfaces of theuntextured surface, surface tex5%, surface tex13% and surfacetex35%, respectively. The micrographs were taken by a high-resolution true color confocal microscope (Axio CSM 700, pro-duced by German Carl Zeiss company). The worn surfaces wereobtained by the wear tests with the parameters listed in Table 3.

Fig. 14(a) shows that the original surface of the untextured speci-men almost worn out. Deep scratches and material loss could beclearly observed, indicating that mixed wear consisting of abrasivewear and adhesive wear occurred during the test. For untexturedsurface, it is difficult to reserve debris and lubricant. On one hand, thedebris generated by friction can causemicro-cutting on the untexturedsurface and lead to abrasive wear. On the other hand, the lubricationwithout back-up lubricant fails easily and the contact area of twomating surfaces increases significantly, resulting in a lot of frictionalheat that is the major cause of adhesive wear. In addition, the lost or

Fig. 14. Optical microscope photographs of worn surfaces with different texture densities. (a) Untextured, (b) tex5%, (c) tex13% and (d) tex35%.

Fig. 13. Effect of texture density on friction coefficient under step increasedrotational speed.

K. Li et al. / Tribology International 77 (2014) 97–105 103

transferred material produced by adhesive wear will scratch thesurface further. Thus, the scratches on the surface become muchwider and deeper.

As shown in Fig. 14(b), abrasive and adhesive wear can also beobserved from the topography of worn surface tex5%, which is similarto that of the untextured surface, while the width and depth ofscratches as well as the area of adhesive wear seem smaller than thoseon the untextured surface. These results suggest that micro-dimplesinfluenced on wear resistance, but the effect was insignificant due tothe limited number of micro-dimples.

For surface tex13%, the increase of texture density improves theability to reserve debris and lubricant. As a result, only severalshallow scratches were observed on the worn surface tex13%, andsome original surface topography can still be observed in Fig. 14(c).

As the texture density increases to 35%, there are more dimplesto reserve lubricant that help cool the friction surfaces andcontribute to the reduction of friction heat generation, this isprobably the main reason for the decrease of adhesive wear forsurface tex35% when compared with surface tex5%, as shown inFig. 14(d) and (b). With the increase of texture density, however,contact area of mating surfaces is reduced simultaneously, givingrise to the contact pressure that causes more severe abrasive andadhesive wear. So when compared with surface tex13%, there aremore deeper scratches on the surface tex35%.

In addition, it is worth noticing that the diameters of the micro-dimples on three textured surfaces (see Fig. 14(b)–(d)) are differ-ent which can also reflect the level of wear. Since the originaldiameter of the dimples before wear test are almost the same, thelarger diameter of the dimple after wear test means relativelyslighter wear. This phenomenon confirms that surface tex13% hasthe best wear resistance in the present study.

In order to further investigate the difference of chemicalcompositions of specimen surfaces after wear tests, the energydispersive spectrometer (EDS) analysis was performed on weartracks of worn surfaces by utilizing the EDS system of scanningelectron microscope (SEM). The EDS results showed that spectra oftextured surfaces with three different texture densities weresimilar. Typical spectra of untextured surface and textured surfacesare shown in Fig. 15(a) and (b), respectively. The weight percentand atomic percent of the elements are listed in Table 4.

From Fig. 15 and Table 4, it is seen that elements of C, O and Cu aredetected on both untextured and textured surfaces. The element withthe highest weight is Cu, which comes from the substrate material ofOFHC copper. The existence of elements C and O is mainly due to theresidue of lubricant oil on worn surfaces. In addition to this, a smallamount of element Fe is also detected on the worn untextured surfaceas shown in Fig. 15(a), while no other elements are found on thetextured surface as shown in Fig. 15(b). Since the hardness of the pin ismuch greater than that of OFHC copper, the material transfer from thepin to OFHC copper surface is difficult to occur. As a result, the transferof the Fe element indicates that wear of the untextured surface ismore severe than the textured surfaces.

4. Conclusions

Micro-dimple arrays with different texture densities werefabricated on OFHC copper plates by laser peen texturing (LPT).The friction and wear performances of textured surfaces anduntextured surfaces were studied in starved lubrication underthe condition of linearly increased normal load and step increasedrotational speed. The following conclusions could be drawn:

(1) It is verified that LPT is an effective technology to fabricatecontrolled micro-dimple array on specimen surface.

(2) Surfaces textured by LPT have better friction performance andhelp to reduce both abrasive and adhesive wear as comparedto untextured surface within the considered range of normalloads and rotational speeds in this research.

(3) Texture density has strong effect on the friction and wearbehavior. It is found that friction coefficient, failure time, loadcarrying capacity and wear resistance do not vary monotoni-cally with the texture density. There might exist an optimaltexture density at which the textured surface exhibits the besttribological performance.

(4) Surface with texture density of 13% shows the best friction andwear performance in this study. Compared with the untexturedsurface, it could reduce friction coefficient up to about 1/2under the test conditions in this research.

Fig. 15. EDS results of worn surfaces. (a) Untextured surface and (b) texturedsurface.

Table 4Composition analysis results of worn surfaces.

Element Untextured surface Textured surface

Weight percent Atomic percent Weight percent Atomic percent

C 23.62 59.66 27.06 64.93O 2.72 5.15 1.47 2.64Fe 0.56 0.30 – –

Cu 73.10 34.89 71.47 32.43

K. Li et al. / Tribology International 77 (2014) 97–105104

Our future work will focus on the qualitative and quantitativecomparison between LPT and LST and further investigate theadvantages and disadvantages of LPT in the application of surfacetexturing.

Acknowledgments

The authors would like to thank the National Natural ScienceFoundation of China (Grant nos. 51075271 and 51375305), Foun-dation for Innovative Research Groups of the National NaturalScience Foundation of China (Grant no. 51121063) and InnovationFund of the National Commercial Aircraft Manufacturing Engineer-ing Technology Research Center (Grant no. SAMC12-JS-15-025).

References

[1] Hamilton DB, Walowit JA, Allen CM. A theory of lubrication by micro-irregularities. J Basic Eng 1966;88:177–84.

[2] Ryk G, Kligerman Y, Etsion I. Experimental investigation of laser surface texturingfor reciprocating automotive components. Tribol Trans 2002;45:444–9.

[3] Etsion I, Kligerman Y, Halperin G. Analytical and experimental investigation oflaser-textured mechanical seal faces. Tribol Trans 1999;42:511–6.

[4] Etsion I. State of the art in laser surface texturing. Tribol-T ASME 2005;125:248–53.[5] Wakuda M, Yamauchi Y, Kanzaki S, Yasuda Y. Effect of surface texturing on

friction reduction between ceramic and steel materials under lubricatedsliding contact. Wear 2003;254:356–63.

[6] Andersson P, Koskinen J, Varjus S, Gerbig Y, Haefke H, Georgiou S, et al.Microlubrication effect by laser-textured steel surfaces. Wear 2007;262:369–79.

[7] Etsion I. Improving tribological performance of mechanical components bylaser surface texturing. Tribol Lett 2004;17:733–7.

[8] Galda L, Pawlus P, Sep J. Dimples shape and distribution effect on character-istics of Stribeck curve. Tribol Int 2009;42:1505–12.

[9] Vrbka M, Šamánek O, Šperka P, Návrat T, Křupka I, Hartl M. Effect of surfacetexturing on rolling contact fatigue within mixed lubricated non-conformalrolling/sliding contacts. Tribol Int 2010;43:1457–65.

[10] Garrido AH, González R, Cadenas M, Battez AH. Tribological behavior of laser-textured NiCrBSi coatings. Wear 2011;271:925–33.

[11] Tang W, Zhou Y, Zhu H, Yang H. The effect of surface texturing on reducing thefriction and wear of steel under lubricated sliding contact. Appl Surf Sci2013;273:199–204.

[12] Nakatsuji T, Mori A. The tribological effect of mechanically produced micro-dents by a microdiamond pyramid on medium carbon steel surfaces in rollingsliding contact. Mechanica 2002;66:663–74.

[13] Pettersson U, Jacobson S. Friction and wear properties of microtextured DLCcoated surfaces in boundary lubricated sliding. Tribol Lett 2004;17:553–9.

[14] Bulatov VP, Krasny VA, Schneider YG. Basics of machining methods to yieldwear- and fretting-resistive surfaces, having regular roughness patterns. Wear1997;208:132–7.

[15] Zhang YD, Lin JQ, Fu QL, Hu HY. Measuring and controlling of arbordisplacement in low frequency vibration machining surface for micropits.JZUS-A 2008;42:1410–4.

[16] Friedrich CR. Micromechanical machining of high aspect ratio prototypes.Microsyst Technol 2002;8:343–7.

[17] Zhou L, Kato K, Umehara N, Miyake Y. Nanometre scale island-type texturewith controllable height and area ratio formed by ion-beam etching on hard-disk head sliders. Nanotechnology 1999;10:363.

[18] Cui FZ, Luo ZS. Biomaterials modification by ion-beam processing. Surf CoatTechnol 1999;112:278–85.

[19] Wang XL, Kato K. Improving the anti-seizure ability of SiC seal in water withRIE texturing. Tribol Lett 2003;14:275–80.

[20] Wang XL, Adachi K, Otsuka K, Kato K. Optimization of the surface texture forsilicon carbide sliding in water. Appl Surf Sci 2006;253:1282–6.

[21] Stephens LS, Siripuram R, Hyden M, McCartt B. Deterministic micro asperitieson bearings and seals using a modified LIGA process. J Eng Gas Turbines Power(Trans ASME) 2004;126:147–54.

[22] Pettersson U, Jacobson S. Influence of surface texture on boundary lubricatedsliding contacts. Tribol Int 2003;36:857–64.

[23] Shinkarenko A, Kligerman Y, Etsion I. The effect of surface texturing in softelasto-hydrodynamic lubrication. Tribol Int 2009;42:284–92.

[24] Ryk G, Etsion I. Testing piston rings with partial laser surface texturing forfriction reduction. Wear 2006;261:792–6.

[25] Etsion I, Halperin G, Becker E. The effect of various surface treatments onpiston pin scuffing resistance. Wear 2006;261:785–91.

[26] Etsion I, Burstein L. A model for mechanical seals with regular micro-surfacestructure. Tribol Trans 1996;39:677–83.

[27] Yu XQ, He S, Cai RL. Frictional characteristics of mechanical seals with a laser-textured seal surface. J Mater Process Technol 2002;129:463–6.

[28] Feldman Y, Kligerman Y, Etsion I. A hydrostatic laser surface textured gas seal.Tribol Lett 2006;22:21–8.

[29] Brizmer V, Kligerman Y. A laser surface textured journal bearing. J Tribol2012;134:031702.

[30] Brizmer V, Kligerman Y, Etsion I. A laser surface textured parallel thrustbearing. Tribol Trans 2003;46:397–403.

[31] Shinkarenko A. Kligerman Y. EtsionI I. Partial elastomer texturing in soft elastohydrodynamic lubrication, In: Proceedings of the STLE/ASME internationaljoint tribology conference; 2008. p. 287–9.

[32] Shinkarenko A, Kligerman Y, Etsion I. The effect of elastomer surface texturingin soft elasto-hydrodynamic lubrication. Tribol Lett 2009;36:95–103.

[33] Iordanova I, Antonov V, Gurkovsky S. Changes of microstructure and mechan-ical properties of cold-rolled low carbon-steel due to its surface treatment byNd: glass pulsed laser. Surf Coat Technol 2002;153:267–75.

[34] Guo YB, Caslaru R. Fabrication and characterization of micro dent arraysproduced by laser shock peening on titanium Ti—6Al—4V surfaces. J MaterProcess Technol 2011;211:729–36.

[35] Kovalchenko A, Ajayi O, Erdemir A, Fenske G. Friction and wear behavior of lasertextured surface under lubricated initial point contact. Wear 2011;271:1719–25.

[36] Montross CS, Wei T, Ye L, Clark G, Mai YW. Laser shock processing and itseffects on microstructure and properties of metal alloys: a review. Int J Fatigue2002;24:1021–36.

[37] Nalla RK, Altenberger I, Noster U, Liu GY, Scholtes B, Ritchie RO. On theinfluence of mechanical surface treatments-deep rolling and laser shockpeening—on the fatigue behavior of Ti—6Al—4V at ambient and elevatedtemperatures. Mater Sci Eng A 2003;355:216–30.

[38] Tan Y, Wu G, Yang JM, Pan T. Laser shock peening on fatigue crack growthbehavior of aluminum alloy. Fatigue Fract Eng Mater Struct 2004;27:649–56.

[39] Wen SZ, Huang P. Principles of tribology. 4th ed.. Beijing: Tsinghua UniversityPress; 2012.

[40] Szeri AZ. Fluid film lubrication: theory and design. 1st ed.. Cambridge:Cambridge University Press; 2005.

K. Li et al. / Tribology International 77 (2014) 97–105 105