Ž .Wear 240 2000 144–151www.elsevier.comrlocaterwear

Comparative wear in titanium diboride coatings on steel using highenergy density processes

Arvind Agarwal, Narendra B. Dahotre)

Department of Materials Science and Engineering, Center for Laser Applications, UniÕersity of Tennessee Space Institute, B.H. Goethert Parkway,Tullahoma, TN 37388, USA

Received 11 October 1999; received in revised form 15 February 2000; accepted 15 February 2000

Abstract

Ž .A comparison between the tribological properties of titanium diboride TiB deposited using high energy density processes such as2Ž . Ž .Pulse Electrode Surfacing PES and Laser Surface Engineering LSE has been made. The wear resistance of TiB coated surface is2

higher than AISI 1010 steel. The wear resistance of the LSE coated TiB coating is even better than that of the PES deposited TiB2 2Ž . Ž .coating. Coefficient of friction values for LSE coated TiB coating ms0.6 are lower than PES deposited TiB coating ms0.7 .2 2

Wear occurs in PES deposited TiB coating by brittle fracture and attrition type mechanisms whereas mixed adhesive–abrasive wear in2

LSE deposited TiB coating occurs by localized plastic deformation of the soft matrix phase Fe from a ‘‘composite’’ layer on the surface.2

q 2000 Elsevier Science S.A. All rights reserved.

Ž . Ž .Keywords: TiB coating; Pulse Electrode Surfacing PES ; Laser Surface Engineering LSE ; Wear resistance; Coefficient of friction2

1. Introduction

Tribological properties of ceramic coatings depend uponseveral factors such as hardness and thickness of thecoating, coating material, type of the substrate, operatingconditions and coating process. Based on these variables,several wear mechanisms have been proposed for differentcoating-substrate system, viz. ploughing, flake formation,micropolishing, cohesive failure or fragmentation, spalling,plastic deformation, removal of hard particle phase and

w xattrition 1 . In the present study, wear and friction charac-Ž .teristics of titanium diboride TiB coating has been2

investigated.TiB is a transition metal based refractory ceramic,2

which has unique properties of extremely high hardness,high melting point, wear resistance, corrosion resistance

w xand electrical conductivity 2 . It also has excellent corro-sion resistance against metallic aluminum and cryolite,which makes it suitable candidate as a coating material forinert cathode used in Hall–Heroult cell for aluminum

w xelectrolysis 3,4 . Conventionally, TiB ceramics are pro-2

) Corresponding author. Tel.: q1-931-393-7495; fax: q1-931-454-2271.

Ž .E-mail address: ndahotre@utsi.edu N.B. Dahotre .

cessed by hot-pressing and sintering which are very expen-w xsive, as it requires high temperature processing 5 . More-

over, near net shape forming is also impossible. However,surface engineering seems to be an alternate way to mod-ify the surface properties similar to that of TiB without2

altering the bulk properties of the engineering componentor structure. In the present study, TiB has been deposited2

on plain carbon steel using high energy density processes,Ž .Pulse Electrode Surfacing PES and Laser Surface Engi-

Ž .neering LSE techniques. A comparative study of wearand friction behavior of TiB coating deposited by these2

two processes is presented. For a better understanding ofthe comparative wear mechanisms, it is essential to under-stand the main features of the two coating processes. Thedetailed comparison of the two coating processes and theresulting coating morphology has been presented else-

w xwhere 6–9 .

1.1. PES

PES technique utilizes a high current, short durationpulse via a discharge capacitance and voltage circuit,which results in melting the electrode and depositing the

w xmaterial on the substrate 10–13 . This process is verysimilar to arc-welding process with the exception of little

0043-1648r00r$ - see front matter q 2000 Elsevier Science S.A. All rights reserved.Ž .PII: S0043-1648 00 00357-4

( )A. Agarwal, N.B. DahotrerWear 240 2000 144–151 145

total heat input. Hence, PES process is also referred asmicro-welding process. The main advantage of this processstems from the fact that metallurgically bonded coatingswith an adherent interface can be produced with little heat

w xinput to the substrate at ambient temperature 14 . Sincethe bulk material remains near ambient temperature andacts as heat sink, heat is rapidly dissipated leading to rapidsolidification of the molten pool which in turn results in anextremely fine grained coating of high density, hardnessand strength. Furthermore, PES process obviates the needof expensive environmental chambers required in severaltypes of other coating processes. The primary requirementand the limitation for the PES process are that electrodematerial should be electrically conductive and thus capableof melting in an electric arc.

1.2. LSE

LSE is one of the surface modification processes wherepreplaced ceramic powder precursor is melted along with athin layer of the substrate to produce a laser melt zone of

w xdesired composition 8 . The most important advantagestems from the fact that laser surface modification is anon-equilibrium synthesis. Laser surface modification in-

Ž 3 8 .volves high cooling rates 10 to 10 Krs , which pro-duces metastable phases by exceeding the solid solubility

w xlimit beyond equilibrium phase diagram 15 . This leads todevelopment of wide variety of microstructures with novelproperties, which cannot be produced by any conventionalprocessing technique. Moreover, these coatings are metal-lurgically bonded providing a sound and adherent interfacebetween the coating and substrate. The surface structurecan be tailored to the surface requirement of the applica-tion by varying process variables such as laser traversespeed, power, beam size and type and composition of the

w xprecursor 16 . Also, laser beam has an excellent spatialresolution that makes it ideal for depositing coating onminiature size components such as electronic sensors forhigh temperature applications. Another advantage of lasersurface modification comes from the fact that laser beamcan be transported to any remote location through fiberoptics. This allows deposition of coatings on componentsor parts that are complex in shape and remotely located.

2. Experimental procedures

2.1. PES coating procedure

Coupons of AISI 1018 steel of dimensions 25 mm=10mm=3 mm were mechanically polished on emery paperof grit 240 and then rinsed with acetone. A sinteredelectrode of TiB containing Ni in the range of 1–3 wt.%2

as binder was used to deposit coating on these steelcoupons. Deposition was done using a hand held gun in airat room temperature. Pulsed electrode deposition was car-

ried out at 50 V and spark duration of 10 ms. Thedischarge capacitance used for the PES process was 450mF with a current of 50 A. Such processing parametersprovide an input energy of 0.10 MWrm2.

2.2. LSE coating procedure

Commercially available TiB and Ti powders obtained2

from CERAC, Milwaukee, WI were used as precursormaterial for this study. Both powders had a purity of99.5%. The average particle size range was 5.5"1 mm.Thin plates of AISI 1010 steel of size 75 mm=150mm=3 mm were cleaned using sand blasting. The precur-sor of a powder mixture of 2 wt.% Ti and 98 wt.% TiB2

composition in a water-based organic vehicle was sprayedon these coupons using a paint spray gun. The averagepreplaced powder bed had a thickness of 150"15 mm.Sprayed coupons were dried at 708C for 1 h prior to laserprocessing. The addition of 2 wt.% Ti to TiB was made2

on the assumption that Ti increases the wettability in thew xmelt zone 17 . It was also anticipated that additional Ti

would recombine with any boron, which may disassociateduring laser processing. A 2.5 kW Hobart HLP 3000

Ž .continuous wave Nd:YAG laser 1.06 mm wavelengthequipped with a fiber optic beam delivery system wasemployed for synthesis of laser assisted TiB coatings. In2

Žorder to provide a large sweeping coverage i.e. rapid.processing speed and to reduce andror to eliminate over-

lap between the successive laser passes, laser optics wereconfigured to provide a 3.5-mm-long line beam in spatial

Ž .distribution TEM onto the sample surface. The process-01

ing parameters were optimized in accordance with earlierw xreported literature 8,16 . In the present study, laser beam

power and laser traverse speed were kept constant at 1.5kW and 33 mmrs, respectively. Thus, the energy densityinput for LSE is 15 MWrm2, which is couple of orders ofmagnitude higher than PES process.

2.3. Microstructural characterization

Coating morphology and wear surface topography wasstudied using a ISI Super III-A scanning electron micro-

Ž .scope SEM coupled with EDAX. Metallographic samplesfor scanning electron microscopy were prepared by polish-ing on Buehler TEXMET 2000 cloth during initial stagesof polishing. The etchant used for TiB coated steel sam-2

ple was Nital. Laser-treated samples required longer etch-ing time as compared to PES samples. Such behaviorcould be explained by the nature of the coatings obtainedin the two processes. Structural characterization for

Ž .phase s identification was carried out on a Philips Norelco˚Ž .X-ray diffractometer with CuK radiation 1.54 A operat-a

ing at 40 kV and 20 mA. Surface roughness of the wearsurface was examined using a WYKO VISION RST 500optical profilometer.

( )A. Agarwal, N.B. DahotrerWear 240 2000 144–151146

2.4. Tribological characterization

Tribological properties of the TiB coating were mea-2

sured using a block-on-disk tribometer. Coated coupons ofdimension 25 mm=25 mm=3 mm were tested for drysliding wear against a hardened steel ring rotating at alinear speed of 4.4 mrs. Weight loss measurements weremade after successive 2 min. The dry sliding wear test wasconducted for a total of 20-min duration with an applied

Ž .normal load of 2 kg. Coefficient of friction m was alsorecorded for a 10-min period by an interfaced computer,which acquired data in the form of electrical output powerof the motor.

3. Results and discussion

3.1. Coating morphology

Coating morphology and surface roughness play a cru-cial role in influencing the tribological properties of the

Ž .coating. Fig. 1 a shows the topographical features of thetop surface of TiB deposited on the steel substrate using2

PES technique. The surface is rough and irregular withseveral islands of globular shape and a splash appearance.This is a characteristic feature of the PES technique. In thepresent study, PES process was conducted in air, whichsupports globular mass transfer mechanism. This is at-

Ž .Fig. 1. a SEM micrograph showing the top surface morphology of theŽ .TiB coated surface using PES process and b the corresponding surface2

roughness profile for the coated surface.

Ž .Fig. 2. a SEM micrograph showing the top surface morphology of theŽ .TiB coated surface using LSE process and b the corresponding surface2

roughness profile for the coated surface.

tributed to the fact that dissociable gases like nitrogen andair form a plasma of high thermal conductivity, which

w xpromotes globular mass transfer 13 . A molten globulardroplet forms at the TiB electrode tip and impinges on the2

steel substrate. This produces a splash appearance on thetop surface. The corresponding surface profile of the coated

Ž .surface is illustrated in Fig. 1 b . The surface is very roughin nature with several peaks and valleys. The average

Ž .surface roughness, R , for PES coating is high 16.91 mma

in tune with its topographical features observed in Fig. 1.The surface topography of the laser-engineered surface is

Ž .shown in Fig. 2 a . The surface is relatively smooth, flatand free from cracks. The profile of the LSE deposited

Ž .TiB surface is illustrated in Fig. 2 b . The average surface2Ž .roughness, R , for LSE coating is low 6.39 mm ina

comparison to PES surface. The importance of the surfaceroughness in wear phenomena will be elucidated in thefollowing sections. Fig. 3 presents the SEM micrographsof the cross-sectional views of the TiB coated steel using2

Ž .PES and LSE processes. It can be seen in Fig. 3 a thatPES deposited TiB coating is homogeneous, dense, ad-2

herent and free of defects like cracks. However, the coat-ing thickness is not very uniform throughout the cross-sec-tion. This is in accordance with the nature of PES processas described above and as earlier observed in the topo-

( )A. Agarwal, N.B. DahotrerWear 240 2000 144–151 147

Ž .Fig. 3. SEM micrographs of the cross-sectional views of a PES sampleŽ .and b LSE sample.

Ž .graphical features of the same sample in Fig. 1. Fig. 3 b isa continuous, adherent and crack free laser deposited TiB2

coating on AISI 1010 steel substrate. The coatingrlaser-melt-zone is ‘‘composite’’ in nature. It is a mixture ofuniformly distributed TiB ceramic particles of different2

Ž .morphologies polygonal and acicular and Fe. During

laser processing, a thin layer of steel substrate melts andmixes with TiB due to convection currents within the2

melt zone resulting in a composite coating. Within thelaser melt zone, the distribution of TiB particles is uni-2

Žform and their content is very high volume fractionf.0.67 .

3.2. Tribological properties of TiB coating2

3.2.1. Weight loss analysisWear test results are presented in Fig. 4, which shows

the weight loss over a 20-min period. It can be easilyobserved that LSE coated surface shows least weight lossin comparison to the PES coated surface and AISI 1010

Ž .steel substrate. The wear rate grmin of the PES coatedsurface has reduced by almost a factor of 3 whereas forLSE surface, wear rate has decreased by a factor of 15 incomparison to the uncoated AISI 1010 steel substrate.Moreover, the wear rate for LSE coated surface tends tostabilize after an initial period of 4 min, suggesting thesmoothening andror breaking off of hard asperities,thereby providing a flat contact with the ‘‘composite’’composition. The worn surfaces of PES and LSE coatedsamples were subjected to topographic observations in

Ž .SEM. Fig. 5 a shows a low magnification SEM micro-graph of the worn region in PES deposited TiB sample.2

Elemental X-ray maps corresponding to Ti and Fe distribu-Ž . Ž . Ž .tion in Fig. 5 a are presented in Fig. 5 b and c , respec-

tively. The wear mechanism in PES deposited TiB coat-2

ing is essentially brittle in nature. Because of the inherentbrittleness of the ceramic material, wear can occur bychipping. Surface and subsurface cracks form, join and

Fig. 4. Cumulative weight losses over a 20-min period during dry sliding wear block-on-disk test. Inset shows wear rates corresponding to various surfaces.

( )A. Agarwal, N.B. DahotrerWear 240 2000 144–151148

Ž .Fig. 5. a SEM micrograph of the worn surface of TiB coated PES2Ž . Ž .sample and corresponding X-ray elemental map for b Ti and c Fe.

release small chips of material. These chipped off particlesare further crushed into finer particles leading to formation

w xof wear debris 18 . Such wear mechanism has been termedas attrition type of wear. A combination of brittle fractureand attrition wear mechanism is observed in PES coatedTiB sample. As observed earlier, PES surface is very2

rough in nature, showing peaks and valleys in the TiB2Ž .coating Fig. 1 . Under sliding conditions of wear, peaks of

TiB coating act as asperities and regions of high stress2

concentration. The interfacial strength between two succes-sive droplets of a refractory ceramic like TiB is not high2

and prone to crack formation. These small cracks tend to

Fig. 6. SEM micrograph showing the crack formation and propagationduring the brittle fracture of TiB -coated PES sample.2

join, leading to the formation of a major crack. Suchphenomenon is clearly evident in Fig. 6, which showssmall and thin cracks merging together to form a widecrack. Such wide crack formation results in chipping ofTiB coating as particles from the steel substrate. Chipping2

of TiB particles suggest a high wear rate, which is2Ž .evident in Fig. 4. Fig. 5 a shows denuded steel substrate

after the chipping of particles of TiB coating. Some of2

these chipped off particles are further crushed and formwear debris, which is indicated by the presence of Ti over

Ž Ž ..the denuded steel substrate Fig. 5 b , while bare regionswith loss of TiB coating on steel substrate are also visible2Ž Ž ..Fig. 5 c . Chipping of TiB particles is further corrob-2

orated by the surface profile of the worn surface shown inFig. 7. TiB coating have chipped off at several locations,2

which is indicated by the deep craters formed on thesurface. The average surface roughness, R , of the worna

Ž .surface reduced from 16.91 mm as coated to 6.27 mm,which is attributed to the flattening of the surface aftersliding wear during the 20-min period.

The wear phenomenon in LSE coated sample is signifi-cantly different from the PES coated sample. As statedearlier, LSE deposited TiB coating is ‘‘composite’’ in2

Fig. 7. Surface roughness profile of the wear surface for the PES sample.

( )A. Agarwal, N.B. DahotrerWear 240 2000 144–151 149

nature. Wear mechanism in such multiphase compositematerials is very complex, depending upon several factorssuch as volume fraction, distribution and morphology of

w xthe ceramic particles 19 . Under ideal conditions, mono-lithic ceramic such as TiB is expected to fail in a brittle2

w xmanner 2 . However, the presence of the softer matrixphase Fe between hard ceramic particles alters the natureof wear mechanism. It has been mentioned earlier that Feacts as an excellent binder for Ti-based refractory ceramicsw x7,20 . This prevents debonding at TiB particlerFe ma-2

w xtrix interface 9 . Under such conditions, Fe-based matrixŽ .phase s deforms plastically to accommodate the high

Ž .Fig. 8. a SEM micrograph of the worn surface of TiB coated LSE2Ž . Ž .sample and corresponding X-ray elemental map for b Ti and c Fe.

Fig. 9. Surface roughness profile of the wear surface for the LSE sample.

stresses experienced by TiB particles. This prevents brit-2

tle fracture and fragmentation of TiB ceramic particles.2

Hence, chipping is prevented and no loose debris is ob-served. A composite layer containing hard TiB particles2

Ž .and soft Fe phase s covers the entire substrate, whichassists in reducing the wear rate. Such microstructuralfeatures are elucidated in Fig. 8, showing TiB particles2

embedded in Fe matrix. There is no crack formationobserved. Wear occurs gradually by mixed adhesive–abra-

Ž . Ž .Fig. 10. X-ray diffraction spectra before B and after A wear ofŽ . Ž .TiB -deposited a PES and b LSE steel coupons.2

( )A. Agarwal, N.B. DahotrerWear 240 2000 144–151150

Fig. 11. Variation in coefficient of friction with time during dry slidingtest for PES sample.

Fig. 12. Variation in coefficient of friction with time during dry slidingtest for LSE sample.

sive mode due to the localized plastic deformation of theŽ .softer matrix, Fe Fig. 4 . Fig. 9 is the surface roughness

profile of the worn surface of the LSE coated sample. Thewear surface clearly exhibits the depressions or gradual

Ž .wear plastic deformation caused by the rotating disk ofthe tribometer and relatively small change in the average

Ž .surface roughness R from 6.39 mm as coated to 4.56a

mm.The X-ray diffraction spectrum of the LSE coated sur-

faces before and after wear is shown in Fig. 10. Within thelimits of resolution of the X-ray diffractometer, the spec-

Ž .trum does not indicate evolution of any new phase s

during the wear process, indicating that wear in LSEcoated sample is essentially physical in nature. Such new

Ž .phase s , if formed, could influence the wear behavior ofthe coatings.

3.2.1.1. Coefficient of friction measurement. In the presentwork, the coefficient of friction is calculated by measuringthe changes in voltage and current in the electrical circuitof the motor driving the block-on-ring tribometer during

w xloading 21 . Based on the principle of energy conserva-Ž .tion, the frictional energy W equals the change in thef

Ž .electrical work during loading, given by Eq. 1 :

W sVoltage DV =Current D I 1Ž . Ž . Ž .f

From the friction theory, it is known that

W smNÕ 2Ž .f

where N is normal load, Õ is the linear speed of the diskand m is the coefficient of friction. By equating the abovetwo equations, the coefficient of friction is computed.

The computed coefficient of friction for TiB coating2Žhas been plotted for the entire test time of 10 min Figs. 11

.and 12 . For PES process, the fluctuations in coefficient offriction are higher in comparison to the LSE process. Thebest-fit line shows coefficient of friction to be about 0.7for PES process and 0.6 for LSE process. Such differencein the friction behavior can be attributed to the coatingmorphology and prevailing wear mechanisms in two coat-ings. In PES process, the surface is inherently rough withpeaks and valleys due to the nature of the process yieldinga higher friction value. Moreover, fluctuations in coeffi-cient of friction values for PES process are higher, insinu-

Ž .ating chipping off TiB particles Fig. 11 . Such fluctua-2

tions are larger for the first 2 min of the test. The degree offluctuation decreases with time, which is attributed to therelative smoothening of the surface. In LSE process, thecoated surface is inherently smoother than the PES surface,which may explain the lower coefficient of friction valueŽ .Fig. 12 . The lower degree of fluctuations in friction valuefor LSE coated sample is accounted by the formation of asmooth ‘‘composite’’ layer on the entire surface. A com-parison between the coatings produced by two processes

Table 1Summary of the TiB coatings and their tribological characteristics deposited by PES and LSE processes2

Features PES process LSE process2Ž .Energy density input MWrm 0.10 15.0

Ž . Ž .Nature of the coating Thin 35–40 mm , homogeneous, Thick 200 mm CompositeŽ .dense and very fine grained 2–3 mm Coating with embedded TiB2

particles in the Fe-rich matrixŽ .Surface roughness, R before wear 16.91 mm 6.39 mmaŽ .Surface roughness, R after wear 6.27 mm 4.56 mma

Ž .Wear rate grmin 0.0005 0.00005Coefficient of friction 0.7 0.6Wear mechanism Brittle failure in addition to attrition Gradual wear in mixed adhesive–abrasive mode

( )A. Agarwal, N.B. DahotrerWear 240 2000 144–151 151

and their tribological characteristics has been summarizedin Table 1.

The wear behavior of the ‘‘composite’’ coating is verycomplex, which could be explained by a model that eluci-dates frictional behavior of multiphase andror composite

w xmaterials 19 . This model predicts that friction behavior ofa compositermultiphase material is influenced by the rela-tive amounts and friction coefficients of the constituentphases. Also, it is affected by the wear resistance of theindividual constituent. In most wear situations, the more

Ž .wear resistant phase s carry a disproportionately largerpart of the load and hence have a larger influence on thefriction of the composite. Hence, the friction coefficient ofthe composite generally does not follow a linear rule ofmixture, but is dominated by the friction coefficient of themost wear resistant constituent. It is well understood that arefractory ceramic like TiB has a high wear resistance as2

compared to plain carbon steel. Hence, according to themodel discussed above, TiB particles in the coating tend2

to carry a larger load than Fe and contribute largely to thecoefficient of friction of the ‘‘composite’’ coating. How-ever, in the present study, no such mathematical modelinghas been done for this particular materials system.

4. Conclusions

1. The wear resistance of TiB coated surface is higher2

than of AISI 1010 steel. The wear resistance of the LSEcoated TiB coating is even better than PES deposited2

TiB coating.2

2. The coefficient of friction values for LSE coated TiB2Ž .coating ms0.59 is lower than for PES deposited

Ž .TiB coating ms0.68 .2

3. Wear occurs in PES deposited TiB coating by brittle2

fracture and attrition type failure mechanisms.4. Wear in LSE deposited TiB coating occurs by plastic2

Ždeformation and gradual wear mixed adhesive-abrasive.mode of the soft matrix phase Fe from a ‘‘composite’’

layer on the surface.

Acknowledgements

The authors wish to express their sincere appreciation toDr. T. S. Sudarshan at Materials Modifications, Fairfaxwho provided the PES samples for the present work. Thesupport extended by Mr. R. Raman and Mr. S. Kru-pashankara at MMI, in depositing coating by the PES

technique is highly acknowledged. The authors acknowl-Žedge the financial support from the US Air Force Contract

.No. F40600-96-C-0004 for this work. The authors wouldlike to thank Ms. Lalitha Reddy Katipelli for her assistancewith image processing.

References

w x Ž .1 S. Bahadur, C.N. Yang, Wear 196 1996 156.w x2 R. Cutler, in: Engineering Materials Handbook: Ceramic and Glasses

Vol. 4 ASM International, Ohio, 1995, p. 787.w x Ž . Ž .3 K. Billehaug, H. Oye, Alum. Verlag Dusseldorf 56 1980 642.w x Ž . Ž .4 H. Oye, M. Sorlie, Alum. Verlag Dusseldorf 2 1994 147.w x Ž . Ž .5 C. Mroz, Ceram. Bull. 78 6 1999 77.w x Ž . Ž .6 A. Agarwal, N.B. Dahotre, Mater. Char. 42 1 1999 31.w x Ž . Ž .7 A. Agarwal, N.B. Dahotre, Surf. Coat Technol. 106 2r3 1998

242.w x8 A. Agarwal, N.B. Dahotre, Int. J. Refractory Mater. Hard Met. 17

Ž . Ž .4 1999 283.w x Ž . Ž .9 A. Agarwal, N.B. Dahotre, L.F. Allard, Pract. Metall. 36 5 1999

250.w x10 R.A. Alfintseva, I.G. Bodrova, A.D. Verkhoturov, J. Less Common

Ž .Met. 67 1979 443.w x11 A.I. Bushik, V.A. Shilov, Elektronnaya Obrabotka Materialov 1

Ž .1991 22.w x12 G.V. Samsonov, A.D. Verkhoturov, V.S. Sychev, A.I. Bezykornov,

Ž . Ž .Fiziko-Khimicheskaya Mekhanika Materialov 7 5 1971 3.w x Ž .13 R.N. Johnson, in: N.B. Dahotre, J.M. Hampkian, J.J. Stiglich Eds. ,

Elevated Temperature Science and Technology I, The Minerals,Metal and Materials Society, 1995, p. 265.

w x14 V.V. Kantsel, T.S. Kurakina, V.S. Potokin, V.I. Rakhovskii, L.G.Ž . Ž .Tkachev, Sov. Phys.-Tech. Phys. 13 6 1968 814.

w x Ž .15 J. Mazumder, in: Kurt H. Stern Ed. , Metallurgical and CeramicProtective Coatings, Chapman and Hall, London, 1996, p. 74.

w x Ž .16 K. Komvopoulos, Materials Science Forum 163–165 1994 417.w x Ž .17 O.M. Akselsen, J. Mater. Sci. 27 1992 1989.w x18 W.A. Glaeser, in: ASM Handbook: Friction, Lubrication and Wear

Technology Vol. 18 ASM International, Ohio, 1998, p. 812.w x Ž . Ž .19 N. Axen, I.M. Hutchings, S. Jacobson, Tribol. Int. 29 6 1996

467.w x Ž . Ž .20 C. Raghunath, M.S. Bhat, P.K. Rohatgi, Scr. Metall. 32 4 1995

577.w x21 P.K. Rohatgi, N.B. Dahotre, Y. Liu, M. Yin, T.L. Barr, in: S.G.

Ž .Fishman, A.K. Dhingra Eds. , Proc. Int. Syposium on Advances inCast Reinforced Metal Composites, ASM International, Ohio, 1988,p. 367.

Arvind Agarwal earned his PhD degree in Materials Science and Engi-neering from the University of Tennessee in 1999. He earned his BTechand MTech degrees in Materials and Metallurgical Engineering from theIndian Institute of Technology, Kanpur, India. He is currently working asa Materials Scientist at Plasma Processes, Huntsville, AL. He is amember of TMS, ASM, American Ceramic Society and AIAA.

Narendra B. Dahotre earned his MS and PhD degrees in MaterialsScience and Engineering from the Michigan State University. He iscurrently a Professor of Materials Science and Engineering at the Univer-sity of Tennessee Space Institute. He is a member of TMS, ASM andSME.