Effect of Microstructure and Thickness on the Friction and Wear Behavior of CrN Coating

on

en

Hard coatings

Thickness

Microstructure

Friction

Wear

ly u

r d

d au

gs

coating thickness. In the present work, the effects of microstructure and coating thickness on the

easingis austrialed wit

[1,2] to forming dies [3,4]. In general, CrN coatings exhibit

stateutionilureating,orse

that the PVD coating microstructure and morphology are depen-

Contents lists available at SciVerse ScienceDirect

.e

Wea

Wear ] (]]]]) ]]]]]]problem has been developed. Plasma enhanced magnetronE-mail address: [email protected] (C. Lorenzo-Martin).superior ductility, fracture toughness, and corrosion and dent on thickness [6]. In addition to deposition parameters, theproperties of coatings, including their tribological properties, willbe inuenced by the microstructure. Recently, a technique cap-able of producing relatively thick CrN lms without the stress

0043-1648/$ - see front matter & 2013 Published by Elsevier B.V.

http://dx.doi.org/10.1016/j.wear.2013.02.005

n Corresponding author. Tel.: 1 630 252 8577; fax: 1 630 252 5568.PleasCrNlike carbon. Of these coatings, CrN is increasingly being used fortribological applications, ranging from automotive components

the adhesion to the substrate. There is often an optimal coatingthickness in terms of failure and performance. It is also knownis usually a suitable coating for a given engineering application.Thin lm coatings of different composition are among the mostcommon ones used for tribological purposes, because they requireminimal to no post-deposition processing. Some examples ofmaterials used for thin lm coatings are TiN, CrN, and diamond-

thickness limitation. The thickness affects the residual stressin PVD coatings and also determines the stress eld distribunder contact conditions, both of which are relevant to fapathways and mechanisms. Usually, the thicker the cothe larger the residual tensile stress accumulation and the wthat produces new coatings with desirable tribological perfor-mance as well as mechanical properties. Currently, many com-mercial coatings are available for tribological applications.Because of the variety of available coatings in terms of thickness,hardness, and fracture toughness, among other properties, there

tion (PVD) techniques, including magnetron sputtering and arcevaporation. The average CrN coating thickness is 15 mm, andthis material exhibits hardness values of HV 14.523.5 GPa and asurface morphology consisting of dense, ne columnar structures.One challenge for PVD coatings (including CrN) is related to the1. Introduction

Thin lm coatings are being incrapplications. Surface engineeringresearch because of the high indcontrol and wear resistance, couple cite this article as: C. Lorenzo-Mcoatings, Wear (2013), http://dx.doifriction and wear behavior of CrN coatings were determined under unidirectional sliding conditions.

Tests were conducted with a dry ball-on-at contact conguration using 1-, 5-, and 10-mm thickcoatings deposited on a hardened H-13 steel substrate by plasma enhanced magnetron sputtering. The

ball specimen was made of WC. The friction behavior was observed to be strongly dependent on coating

thickness and microstructure, especially at relatively low loads (5 N). At higher loads, however, the

thinnest coating (1 mm) was quickly worn through, while the thicker ones (5 and 10 mm) remainedintact. Wear in both the counterface WC ball material and the coatings also depended on coating

thickness and microstructure. In all coatings, there was localized damage but minimal wear. Additional

tests were done with Si3N4 and 52100 steel balls, and the results indicated different wear and friction

behavior from that for WC balls. The observed effect of coating thickness on tribological behavior is

attributed to differences in the microstructure and mechanical behaviors of CrN coatings as a function

of thickness.

& 2013 Published by Elsevier B.V.

ly used for tribologicalfast growing area ofdemands for frictionh enabling technology

oxidation resistance compared to the widely used TiN counter-part. Also the lower coefcients of friction and higher wearresistance under dry sliding conditions make CrN coatings excel-lent candidates for a variety of applications, such as metal-forming and die-casting tools [5].

CrN coatings are produced mainly by physical vapor deposi-Keywords:Effect of microstructure and thicknessbehavior of CrN coatings

C. Lorenzo-Martin a,n, O. Ajayi a, A. Erdemir a, G.R. Fa Argonne National Laboratory, Tribology Section, Lemont, IL 60439, USAb Southwest Research Institute, San Antonio, TX 78228, USA

a r t i c l e i n f o

Article history:

Received 31 August 2012

Received in revised form

4 February 2013

Accepted 7 February 2013

a b s t r a c t

One of the most common

deposited by physical vapo

and forming tools, dies an

engines. In selecting coatin

journal homepage: wwwartin, et al., Effect of micr.org/10.1016/j.wear.2013.02the friction and wear

ske a, R. Wei b

sed tribological thin-lm coatings is chromium nitride (CrN), typically

eposition. Examples of current applications of this coating include cutting

tomotive components, such as injection valves and piston rings for diesel

for different tribological applications, one of the critical parameters is the

lsevier.com/locate/wear

rostructure and thickness on the friction and wear behavior of.005i

sputter (PEMS) deposition is an improved version of conventionalmagnetron sputtering. The PEMS technology can produce CrNcoatings as thick as 20 mm [7]. With the advent of PEMS, it is thuspossible to evaluate the impact of coating thickness on tribologi-cal performance over a broad range of coating thicknesses. Thistype of study was not possible prior to PEMS because the largestthickness achievable, without stress problems, was typically lessthan 3 mm.

In this study, we used a tungsten lament and a discharge powersupply to generate a global plasma in the entire vacuum system andfabricated CrN coatings of different thickness (110 mm). The goalwas to assess the effect of CrN coating thickness and microstructureon tribological performance.

2. Experimental details

2.1. Coating material

CrN coatings of different thicknesses were deposited onpolished H-13 steel at specimens with dimensions of21.50.25 in.3 and hardness of 58.5 Rc by the PEMS process.The H-13 material was hardened by heat treatment throughquenching and tempering at 550 1C. Fig. 1 is a schematic of thePEMS system developed at Southwest Research Institute [6]. Priorto coating deposition, substrates were sputtered with Ar for6090 min to remove residual contaminant on the surface.

Power

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]]2Cr Target

Sam

ple

s

To Pump

Ar, N2

Worktable

Supply

-

Bias Power Supply

-

MagnetronA solid target of 170-mm diameter was employed as a sourceof Chromium (Cr). The power was set at 1.5 kW. During thedeposition process, a mixture of argon (Ar) and nitrogen (N2)gases was introduced, and the chamber pressure was maintainedbetween 0.33 and 0.47 Pa (2.53.5103 Torr). The surfacetemperature of the substrate was maintained at about400 1C [8]. Because this temperature is lower than the temperingtemperature, no change in microstructure and properties isexpected for the substrate during the coating process. Coatingswere deposited to approximately 1-, 5-, and 10-mm thicknesseswith a corresponding surface nish of 0.011, 0.024, and 0.025 mm

Filament

GlobalPlasma

Magnetron

AC

Discharge Power supply

+

+Magnetron-generated PlasmaFig. 1. Schematic of plasma enhanced magnetron sputtering (PEMS).

Please cite this article as: C. Lorenzo-Martin, et al., Effect of micrCrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02Ra, respectively. The roughness of the original polished steel atsurface was about 0.005 mm. Coatings surface morphologies, aswell as cross sections, were examined by using optical andscanning electron microscopy (SEM) (Quanta 400F ESEM) operat-ing at 10 kV.

2.2. Hardness measurement

Coating surface hardness (H) and reduced Youngs modulus(Er) were measured by nanoindentation method using a triboin-denter TI-950 (Hysitron Inc.). A Berkovich diamond tip with anapproximate radius of curvature of 50 nm was used for thehardness measurements. A xed displacement of 100 nm wasselected, and indents were performed at a loading/unloading rateof 20 nm/s. The reduced Youngs modulus was obtained from theslope of the unloading part of the load-displacement curve usingthe Oliver and Pharr method, as described in Ref. [9]. Because theroughness of the coating seemed to be too large to obtainconsistent measurements, indentations were placed on a polishedannular ring of the coating produced by a micro-abrasion ballcratering method to measure the hardness and elastic modulusaccurately close to the coating surface (less than 0.25 mm fromthe top surface). A minimum of six indents were considered for anaverage of hardness and reduced modulus calculation.

The crystalline structure of the coatings was assessed by usinga Philips X Pert X-ray diffractometer. A Cu Ka radiation source(l0.1542 nm) was used, operating at 40 kV and 40 mA. X-rayscans covered from 201 to 1001 at a step size of 0.011 and scanspeed of 0.1251/s. The total number of steps was 8000. Peaks wereidentied by using the X Pert data viewer software.

2.3. Friction and wear testing

Friction and wear tests were conducted in unidirectional drysliding contact using a ball-on-at conguration (Fig. 2). The ballshad 1/2 in. diameter, were made of WC, and had a hardness valueof 16 GPa and a precision surface nishing grade of 25 (38 nmmeasured surface roughness). According to manufacturer speci-cations, the ball material consisted of 9094% tungsten and 610%cobalt as a binder material. The steel substrate and CrN coatingswith three thicknesses were tested against the WC balls. Forcomparison, 52100 steel and Si3N4 balls were also tested.

Tests were conducted at normal loads of 5, 10, and 20 N. Themaximum Hertzian contact pressures calculated for steel at testedagainst WC ball were (0.91, 1.15, and 1.44 GPa, respectively). ForSi3N4 ball tested against steel at, the maximum Hertzian contactpressures were ( 0.77, 0.96, and 1.22 GPa, respectively), while for the52100 steel ball on steel at the maximum Hertzian contactpressures were (0.68, 0.86, and 1.09 GPa, respectively). The effect ofcoating on contact pressure will be dependent on coatings properties(see Table 1). Compared to steel substrate, coatings with lowerelastic modulus than steel (such as CrN-3) will produce lowercontact pressures, while coatings with higher elastic modules thansteel (such as CrN-1 and CrN-2) will produce higher contactpressures than steel substrate. The contribution of the coating tocontact pressure will be also dependent on load and coatingthickness. The higher the load and the thinner the coating, the moreimportant contribution of substrate to contact pressure. All testswere run for 60 min and at a linear speed of 1 cm/s in ambient roomair (relative humidity of 65%). The friction coefcient was continu-ously measured during the sliding. Wear on ats and counterface ballmaterials was evaluated at the conclusion of the test by using opticalprolometry. We obtained 3D maps of the ball scars and measuredvolume wear for each track. To calculate the wear volume of thecoatings, we measured the wear track proles across the sliding

direction of the track at four locations, also using the prolometry

ostructure and thickness on the friction and wear behavior of.005i

spectroscopy (EDS) analysis.

r un

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 33. Results and discussion

3.1. Surface morphologytechnique. Wear rates were obtained by normalizing wear volumeby the total sliding distance. A repeat test was conducted for eachcondition. Worn surfaces were also characterized by optical andscanning electron microscopy equipped with energy dispersive X-ray

Fig. 2. Ball-on-at test rig fo

Table 1Properties of coatings used in this work.

Coatings Deposition

method

Thickness

(mm)Reduced elastic

modulus (GPa)

Hardness

(GPa)

Roughness

(nm)

CrN-1 PEMS 1 256 21 11

CrN-2 PEMS 5 269 21 25

CrN-3 PEMS 10 190 14 24

Steel none 0 212 8.4 5Fig. 3 shows the surface morphology for the three CrN coat-ings. Average grains size (d), as determined by surface imageanalysis, is dependent on coating thickness: d0.1 mm for CrN-1,d0.40.5 mm for CrN-2, and d0.81.2 mm for CrN-3. Althoughgrain size distribution is mostly homogeneous for all the coatings,the surface morphology (specically the aspect ratio) is depen-dent on coating thickness.

For the thinnest coating, the grains are uniaxial, while thethickest coatings possess a clear columnar structure with thegrains preferentially elongated perpendicular to the surface.Cross-sectional micrographs of the coatings are shown in Fig. 4.

For CrN-1, the coating is mostly uniaxial. Over a Cr bond layerof 400 nm, a columnar CrN structure that is tens of nanometersthick and few hundreds of nanometers high is starting to grow.For the thicker coatings, CrN-2 and CrN-3, the structure is clearlycolumnar with cauliower-like texture on the surface (Fig. 5b).The columns grow wider with thickness of the coatings, forming afan shape. The columns are slightly tilted 351 with respect to theperpendicular direction of the substrate. In the thickest coating,CrN-3, the columns grow from a width of 50 nm (close to thesubstrate) to 500750 nm at the thickest area, giving a coarserstructural morphology. Also, surface localized defects (known asmacro particles or macro droplets) were observed for all coatings,especially for thicker ones.

Fig. 5 shows such a typical nodular-like defect. The particledefect diameter is in the 210 mm range. These types of defectsare produced during the deposition process, and because thebonding of the macro particle with the coating is relatively poor,

Please cite this article as: C. Lorenzo-Martin, et al., Effect of micrCrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02they can be easily removed during the wear process, leavingcraters on the coating surface. The presence of macro particlesmay be responsible for the accelerated initial wear.

3.2. Crystal structure of the coatings

X-ray diffraction (XRD) spectra for the coatings are presentedin Fig. 6. These spectra show how the relative peaks for Fe, Cr, andCrN change with coating thickness. The uncoated steel (baseline)shows a clear iron peak at 44.451. For the thinnest coating, CrN-1,the most intense peak is located at 44.521. Two possible peaksoverlap at this position. One peak is Fe-a from the substrate;considering the relative thin coating of less than a micron, theeffect of the substrate can be important.

The second possible overlapping peak is Cr (110), most probablyfrom an initial Cr bond layer during deposition. Another peak almostequally intense for the CrN-1 coating is CrN (111). Several otherweaker peaks, corresponding to Cr and CrN, are also observed. Asthe coating thickness is increased to 5 mm (CrN-2), the CrN (111)peak dominates, followed by CrN (220) and several other weakerpeaks for CrN and Cr. With coating thickness of 10 mm (CrN-3), thestrongest peak is CrN (220), followed by CrN (200). The CrN (111)peak is still present but is less intense than for the thinner coatings,and a new peak corresponding to CrN (311) appears. In sum, theXRD analysis showed that the three coatings consist of CrN crystalsprimarily, regardless of coating thickness.

3.3. Nano-mechanical properties of the coatings

Table 1 summarizes the measured hardness and reducedYoungs modulus obtained for the three CrN coatings, determined

idirectional sliding testing.using the nano-indentation method described earlier in experi-mental details section. Other relevant properties of the coatingand substrate material, such as thickness and roughness, are alsoincluded.

Note that CrN-1 and CrN-2 have very similar hardness(21 GPa) while CrN-3 exhibits a considerably lower hardness(14 GPa). The reason for differences in the measured hardnessvalues is unclear, perhaps a reection of differences in structuralmorphologies and/or differences in residual stresses in the coatings.

3.4. Friction and wear results

The friction behavior of the three CrN coatings and uncoatedsteel substrate tested under unidirectional dry sliding with theWC ball is summarized in Fig. 7. For the uncoated H13 steel at,the friction behavior is nearly identical at 5 and 10 N loads. Inboth cases, the friction coefcient increased rapidly at the start ofthe test to a maximum value of about 0.6, followed by a gradual


ren

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]]4Fig. 3. Surface morphology of CrN coating of diffedecrease to a minimum value of about 0.4, after which acontinuous gradual increase occurred to the end of test. For therst 1000 s of sliding at 20 N (Fig. 7c), a similar trend wasobserved: rapid increase, followed by both gradual sequentialdecrease and increase. However, this trend was followed by asudden rapid decrease to a near constant value of 0.35 for theduration of the test.

This friction behavior was observed in repeat tests for the WC andH13 steel sliding pair. It is attributed to the extensive formation of atransfer lm on the uncoated steel at, as shown in Fig. 8. Higherloads produced more transfer lm. In the 20-N load test, the rapiddecrease in friction coefcient after 1000 s of testing is attributed tolarge coverage of the contact area by the WC transfer lm, such thatin the later stage of testing, the contact interface consisted of thesliding of the WC transfer lm on the WC, which is known to havefriction coefcient of about 0.350.45 [10]. There is also the possibilityof oxidation at the tribo contact interface resulting in the formation ofiron and tungsten oxides. Oxidation of wear debris trapped at thecontact interface can also occur. Indeed, tribo layers are well known

Fig. 4. Cross-section SEM micrographs of CrN coatin

Fig. 5. Characteristic surface defect (m

Please cite this article as: C. Lorenzo-Martin, et al., Effect of micrCrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02t thicknesses: (a) 1 mm, (b) 5 mm, and (c) 10 mm.to consist of a complex mixture of both material pair in contact, aswell as species from the environment.

The friction coefcients in the tests with CrN-coated ats werealmost identical at 5 and 10 N loads. For the three coatings atthese loads, the friction coefcient showed a general gradualincrease with time. For both loads, the friction coefcient (m) inCrN-1 was consistently higher than for the thicker coatings, witha value of m0.7 at the test conclusion. The CrN-2 and CrN-3coatings showed identical friction coefcients for the rst fewhundred seconds of testing, and by the end of test, the frictioncoefcient with CrN-2 was about the same or slightly lower thanthat for CrN-3 (m0.5 vs. 0.55).

At 20 N, the friction behavior in tests with CrN-2 and CrN-3remained unchanged, but a sudden rapid increase in noise wasevident in the test with CrN-1 after 1500 s. This transition in CrN-1 coincided with the coating being worn through, as shown inFig. 9a. It is possible that the sudden increase in friction was dueto the ball sliding against the Cr interlayer, although unlikely.Rather, the friction increase is more likely due to the abrasive

g: (a) 1 mm, (b) 5 mm, and (c) 10 mm thickness.

acro particles) for CrN coatings.


plowing action of the hard coating debris generated by thecoating failure.

Subsequent WC transfer occurred from the ball to the partiallyexposed steel substrate (Fig. 9b), as indicated by EDS analysis(Fig. 9c). This WC transfer may account for the decrease in frictioncoefcient from 0.7 to 0.5 at the very end of the test, consideringthat the new contact interface consists of WC ball material against aat surface with CrN-steel-WC areas. For the thicker coatings (CrN-2and CrN-3), the friction behavior and magnitude of coefcients aresimilar to those in the lower load tests. Although no steady frictioncoefcient value was achieved by the conclusion of testing for mostof the materials and loads tested, an average friction coefcient was

calculated for the duration of the test, excluding the rst fewminutes of testing. Average friction coefcient values are summar-ized in Fig. 10.

The highest average friction coefcients are observed for thethinnest coating (CrN-1), for which friction decreases with load,exhibiting values ranging from 0.64 (for 5 N) to 0.58 (for 20 N). Forthe thickest coatings (CrN-2 and -3), the average friction coefcientsvalues are similar to those of the uncoated steel baseline. While forthe uncoated steel at average friction values decrease with load too,from 0.51 (for 5 N load) to 0.37 (for 20 N load), they do not for theCrN-2 and CrN-3 coatings. Of the coatings CrN-2 exhibits the lowestaverage values: 0.38, 0.48, and 0.44 for 5 N, 10 N, and 20 N,respectively. For CrN-3 the values are slightly higher: 0.44, 0.51,and 0.49 for 5 N, 10 N, and 20 N respectively. While the frictioncoefcients for repeat tests were quiet close (less than 4% variability),more variations were observed in the wear values (up to 10%).

Fig. 11 shows the wear results, as measured by optical prolo-metry, in the ball-and-at specimens after testing. For the test withCrN-1 at highest load (20 N), in which the WC ball counterface hadworn through the coating, the highest wear rates were measured intheWC ball and at. With that exception, wear in both WC balls andats appears to increase nearly linearly with load for the other testpairs. Wear in the uncoated steel at was, on average, at least anorder of magnitude higher than that for the coating material. Coatingthickness seems to have a minor effect on coating wear, with thetwo thickest coatings wearing slightly more than the thinnest one atlow loads (Fig. 11a). The same tendency was observed for ball wear(Fig. 11b). While coating thickness showed minimal effect on WCball wear, the material tested against it had a major effect on wear.

Fig. 7. Friction evolution with time for unidirectional dry sl

20 30 40 50 60 70 80 90 100-50

0

50

100

150

200

250

300

350

2 Theta

Inte

nsity

(co

unts

)

H-13

CrN-1

CrN-2

CrN-3

Fig. 6. XRD peaks of CrN coatings deposited on H-13 steel substrate.

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 5Fig. 8. Steel at tested against WC ball in dry unidirectional sliding: (a) overa

Please cite this article as: C. Lorenzo-Martin, et al., Effect of micrCrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02The WC ball wear against the uncoated steel at was, on average, atleast four times higher than ball wear against CrN coatings. Althoughsteel hardness is considerably lower than CrN coating hardness (seeTable 1), this higher wear is attributed to the presence of relativelyhard second-phase carbide particles, notably VC, MoC, and CrC,sliding against the H-13 steel sample microstructure and surface.

iding at different loads: (a) 5 N, (b) 10 N, and (c) 20 N.ll track, (b) transfer layer from ball, and (c) EDS analysis of transfer layer.


rack

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]]6Fig. 9. CrN-1 coating tested against WC ball in unidirectional sliding: (a) overall tdamage and transfer layer.

0.1

0.2

0.3

0.4

0.5

0.6

0.7

5N10N20N

Ave

rage

Fric

tion

Coe

ffici

entWear on the uncoated steel at was predominantly abrasive wear.The transfer layer of WC from the ball onto the surface protected itpartially against further wear (as shown in Fig. 8). In terms of wearmechanisms in the CrN coatings, polishing wear and some transferfrom the ball material are the dominant mechanisms. The exceptionis CrN-1 tested at high load (20 N). As the coating was worn through,severe abrasive wear became the dominant mechanism (as seen inFig. 9b). By contrast, CrN-2 showed mostly polishing wear withminimal transfer lm formation regardless of the test load. Somelocal damage, consisting of cracking and delamination of the transfersurface layer, was observed (Fig. 12).

For CrN-3, load seemed to affect the degree of damage on thewear track. For low load (5 N), there is minimal damage to theoriginal surface morphology. Only mild polishing and some smooth-ing of the wear track byminimal transfer are apparent. At higher load(20 N), the amount of transfer is increased considerably, and exten-sive patches of WC transfer are observed (Fig. 13b). This transferlayer, which has a smoothing effect on the track, eventually grows toa critical thickness and fails by cracking and chipping (Fig. 13c).

3.5. Effect of counterface material

Additional friction and wear tests were conducted with the CrN-2coated at sliding against polished, commercially available bearingballs composed of either hardened 52100 steel (Ra36 nm) or Si3N4(Ra10 nm). The hardness of the 52100 balls is about 7.2 GPa(62 Rc), and that of the Si3N4 balls is about 14.5 GPa. Tests with theseballs were conducted by the same procedure as used with WC balls.

Fig. 14 shows the friction variation with time during the test withthe three balls (WC, 52100 steel, and Si3N4) when sliding against

0H-13 CrN-1 CrN-2 CrN-3

Fig. 10. Average friction coefcient for CrN coatings and uncoated substratesliding against WC ball.

Please cite this article as: C. Lorenzo-Martin, et al., Effect of micrCrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02CrN-2 at 10 and 20 N loads. For WC, the test started at relatively lowfriction coefcient of 0.2, but increased gradually for the 1-hourduration of the test, ending with a nal value of about 0.5. The rate offriction increase decreased with sliding distance or time. For thismaterial, the friction coefcient appears to be independent of load, asthe tests at 10 and 20 N showed nearly identical friction behavior. Thepresence of the cobalt binder phase in the ball may have contributedto the friction coefcient and noise reduction. For the 52100 steelballs, sliding started with a friction coefcient of about 0.8 at both 10and 20 N load. This was followed by a slight gradual decrease to asteady value of about 0.70 for 20 N and about 0.75 for 10 N. Thefriction coefcient data showed considerably more noise in the testswith the steel ball compared to the WC ball. The high friction andnoise in the test data for the steel ball are attributed to extensivemetal transfer into the coated at surface, as illustrated in Fig. 15.

The tests with Si3N4 balls started with a friction coefcient ofabout 0.25 at both 10 and 20 N. In both cases, the friction coefcientincreased rapidly during the rst 5 min (300 s) of sliding, reaching anearly constant value of 0.7 for the remainder of the one-hour test.Furthermore, the friction coefcient data increased in noise through-out the test. This friction behavior is the result of both transfer ofmaterial from the Si3N4 ball into the coating and the wear anddamage of the coating, as shown in Fig. 16.

Fig. 17 compares the wear for the three different balls with CrN-2coated ats. Wear was the greatest for the steel ball and the least forthe WC ball (Fig. 17a). Although more wear occurred in the Si3N4ball compared to the WC ball, it was substantially lower than thewear in the steel ball. This behavior is perhaps a reection ofdifferences in ball hardness. The steel ball produced no wear interms of material removal from the coated at, while the Si3N4 ballproduced the most wear on the CrN coated at (Fig. 17b). The wearproduced by the WC ball is signicantly less than that for the Si3N4ball due to the more extensive formation of a transfer layer, eventhough the WC ball is harder than the Si3N4 ball. Note that thefriction coefcients in the test with the WC ball are also much lower

, (b) transfer layer from ball and coating damage, and (c) EDS analysis of coatingthatheof w

4.

promic

ostr.00n those in the test with the Si3N4 ball. This condition will reduceshear stresses imposed on the coating and, hence, the amountear.

Summary

Although CrN coatings with different thicknesses can beduced by PEMS, they have signicant differences in terms ofrostructure and surface morphology:

For the 1-mm thick coating the grain structure is uniaxial. Thecrystalline structure is mostly CrN with possible smallamounts of Cr. In terms of morphology, this coating surfaceis relatively smooth.

ucture and thickness on the friction and wear behavior of5i

For the 5-mm thick coating, the grain shape distribution isbimodal. There are uniaxial grains close to the substrate andcolumnar-shaped grains in the upper region of the coating.The crystal structure is mainly CrN. Surface morphology iscoarser with small cauliower texture.

The 10-mm thick coating is mostly columnar with fan shape(wider columns as the coating thickens). Also, the crystalstructure is mainly CrN, and its surface morphology is thecoarsest with cauliower texture.

These differences have a signicant impact on the coatingproperties and tribological behavior. For example, high hardness inthe CrN-1 and CrN-2 coatings is most probably due to the grainmorphology. The CrN-3 has lower hardness due to grain morphologyand/or residual stresses. In unidirectional sliding against the WC ball,friction in the 1-mm coating is higher, while it is similar for the 5- and10-mm coatings. Since CrN-1 is the only coating with signicantlydifferent microstructure, its higher friction coefcient may be due toits different microstructure, including higher level of Cr content.Usually, differences in surface morphologies result in differences in

: (a) overall track, (b) mild polishing, and (c) localized damage.

ishing at low load, (b) extensive transfer at high load, and (c) localized damage in

1.3

0.27 0.260.51

2

0.32 0.45 0.81

4.4

6.5

0.511.1

0

1

2

3

4

5

6

7

8

Steel CrN-1 CrN-2 CrN-3

5N

10N

20N

3.3 0.9 0.2 1.7

15.4

1.5 1.0 3.5

69.9

134.0

0.8 4.30

20

40

60

80

100

120

140

Steel CrN-1 CrN-2 CrN-3

5N

10N

20N

Flat Wear x103 m3/m Ball Wear x105 m3

Fig. 11. Flat-and-ball wear for CrN coatings and uncoated substrate at different loads.

Fig. 14. Friction behavior of CrN-2 coating when sliding against different ballmaterials at 10 and 20 N.

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 7Fig.

tran

PC13. CrN-3 surface after testing in dry sliding against WC ball: (a) mild polFig. 12. CrN-2 surface after testing in dry sliding against WC ballsfer layer.

lease cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior ofrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i

g du

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]]8Fig. 15. (a) Transfer from steel ball to CrN-2 coatininteractions between coating and ball surface asperities. This effectmay result in differences in the amount of material transfer and,hence, in friction and wear behavior. The CrN-2 and CrN-3 samples(thicker coatings) exhibited more local surface defects (macro-parti-cles/macro-droplets). This condition resulted in local surface damagein these coatings during sliding contact. There are other possibledifferences in the structure and microstructure of the three coatingsin this study that may account for some of our observations. Morecomprehensive comparative structural analysis by multiple techni-ques and tools are needed to further elucidate the observationsreported in the present paper.

In general, thicker coatings showed better friction and wearbehavior, while the thinnest coating was easily worn through at highload, resulting in a substantial increase in friction and wear. Com-pared with the uncoated steel surface, the coatings have minimaleffect on friction but signicant effect on wear reduction. Indeed,lower friction was observed in tests with uncoated steel under someconditions after a good transfer layer had formed. This transfer layeron steel at consists of a signicant amount of tungsten. Thus, it ispossible that lower friction in uncoated steel is due to the formationof the W-rich transfer layer [11].

Fig. 16. (a) Transfer from Si3N4 ball to CrN-2 coating d

0.0

1.0

2.0

3.0

4.0

5.0

WC Si3N4 Steel

Wea

r V

olu

me

(105

m

3 )

10N

20N

Fig. 17. Comparison of ball and at wear for tests at 10 and 20 N: (a) CrN-2 coatin

Please cite this article as: C. Lorenzo-Martin, et al., Effect of micrCrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02ring dry sliding; (b) EDS analysis of transfer layer.When sliding against different counterface materials, theCrN-2 coating showed signicant differences in friction and wearbehavior. Although more studies are needed to further elucidatethe role of counterface materials, this preliminary study indicatesthat CrN coatings should be paired carefully with counterfacematerials. The current paper suggests that thicker coatings can bedeposited for tribological applications, but more work is neededto optimize coating thickness for different applications.

Acknowledgments

This work was supported by U.S. Department of Energy,Energy Efciency and Renewable Energy, Ofce of Vehicle Tech-nologies, under contract DE-AC0206CH11357. The electronmicroscopy was accomplished at the EMC at Argonne NationalLaboratory, a U.S. Department of Energy Ofce of ScienceLaboratory operated under Contract No. DE-AC0206CH11357by UChicago Argonne, LLC.

uring dry sliding; (b) EDS analysis of transfer layer.

Wea

r V

olu

me

(105

m

3 ) 356.6

0

100

200

300

400

W C Si3N4 Steel

10N

20N

g wear against different balls and (b) wear of different balls against CrN-2 at.


References

[1] H. Scheerer, H. Hoche, E. Broszeit, C. Berger, Surface and Coatings Technology142144 (2001) 1017.

[2] S.C. Tung, H. Gao, Wear 255 (712) (2003) 1276.[3] J. Vetter, R. Knaup, H. Dweletzki, E. Schneider, S. Vogler, Surface and Coatings

Technology 8687 (1996) 739.[4] B. Navinsek, P. Panjan, Surface and Coatings Technology 7475 (1995) 919.[5] P. Panjan, M. Cekada, R. Kirn, M. Sokovic, Surface and Coatings Technology

180 (2004) 561.

[6] J.A. Thorton, Annual Review of Materials Science 7 (1997) 239.[7] Ronghua Wei, Edward Langa, Christopher Rincon, James H. Arps, Surface and

Coatings Technology 201 (2006) 44534459.[8] Feng Cai, Xiao Huang, Qi Yang, Ronghua Wei, Doug Nagy, Surface and

Coatings Technology 205 (2010) 182188.[9] W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (1992) 1564.[10] K. Bonny, P. De Baets, J. Vleugels, S. Huang, B. Lauwers, Tribology Transac-

tions 52 (2009) 481491.[11] H. Engqvist, H. Hogberg, G.A. Botton, S. Ederyd, N. Axen, Wear 239 (2000)

219228.

C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 9Please cite this article as: C. Lorenzo-Martin, et al., Effect of micrCrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02ostructure and thickness on the friction and wear behavior of.005i

Effect of microstructure and thickness on the friction and wear behavior of CrN coatingsIntroductionExperimental detailsCoating materialHardness measurementFriction and wear testing

Results and discussionSurface morphologyCrystal structure of the coatingsNano-mechanical properties of the coatingsFriction and wear resultsEffect of counterface material

SummaryAcknowledgmentsReferences

Effect of Microstructure and Thickness on the Friction and Wear Behavior of CrN Coating

Documents

Transcript of Effect of Microstructure and Thickness on the Friction and Wear Behavior of CrN Coating