NASA Interdisciplinary Collaboration In Tribology

surfacerelativeTribologiststo another),regardingat leastthe terminologyOften the words "adhesiveWear.changeably,even though itin mild wear."delaminationTribologist.to bringsome unificationregardsthe classifications,Althoughsome of thisexpressedof Friction.thatAn attemptpublication[4],the birdscrystalsof birdsfamilyin the monoclinicasof wear.own viewsalso takes account of theexpressedclassesin theirreviewsofthe variousexactlyshould have at leastcontactthose phenomena bethe seven crystalresistancesystemsof wear in Tribology.system have at leastabout thatand alla group of phenomenahave the same beak structure;lookingconrmon, possiblyof Wear, especiallyin Ornithology,which leaves the crystalclassificationreview,in cormnon, whetheror the variousin the finchaxis;be made, in thisis a means of describingone featurefamiliesin Crystallography,iteven toof WearA "classification"the variouswillobvious,wear",respectively.The Classificationswhich have at least"scuffingis based on the author'sviews of both Tabor [S] and Ludema [6],1.2wear",a very small rolelaws, mechanisms and theoriesunification'Wear" and "Scuffing"wear" are used inter-wear" are not at allto the terminologyin a previousThis is not so foradhesion playsbetween "severewear" and "adhesivethe experiencedis agreement amongstwear" and "mildis clearThe differencestheretheone axis of rotationthe same formanifestationsone observableor finalallAllsurface180" rotationof a given wearcharacteristic,topography.inQuite

often,in tionsor theoreticalClassificationdescriptionare made where thereknowledge to propose laws oris the firstthere have recentlyappeared some new classificationsto the type of wear particlethey are stronglyrelatedAs willwhilstattemptbe seen later,k-chardand HirstBurwelland HirstTabor [S] and the authora more viableEssentially,classificationsis cleara definitive[4] have proposed similaris a distillationsystem of classificationor Archard and HirstItof theirthan eitherthe author[4] maintainsthatBurwellnot always appreciatedthatresistancethese classificationssize and compositionwear theory.toideas as regardsand Strang[8](ii)analysissurfaces[9]and easy toand "severe",itisare based on (i) measure-thatof the wear debris(severe wear normallyand Hirstphenomenological.(severe wear is characterizedwhereas mild wear providesresistance);systems,the Archardsince they are entirelyBecause of the every day usage of the words 'mild"contactsomemodificationsof mild and severe wear are the most basicments of contactthat[9].apply to any wear situation,resistanceseven wear classesthese two classificationstep to formulatingThe followingthese systems.proposed by Burwelland Strang envisagedenvisaged only two.preliminary[7]),[9] in 1952 and 1956 respectively.should be made to reconcileas a necessary(Bowden and Westcottto the classifications[8] and Archardstep towards theof a phenomenon.of wear accordingand Strangisconsistsby low contactgive mainlyhighboth as regardsof large, lO -2 mm

diameter,metallicparticles,whereas mild wear debrisdiameter,and has been produced partiallyatmosphere or fluid);and (iii)(severe wear leavesthe surfacesproduces extremelysurfacefinish).specifyations,itis possibleto the transitionper unitslidingmild wearmuch smoother than the originalthese classificationsare oftendo notUnder certainto occur at a ratefor the same combinationof materials.two ordersfrom the slidingis itselfIn general,of magnitude moresurfaces,slightlyespeciallymisleading.it means the volume of materialdistanceof thatwhich the relativemotionsurfacewithclose[lo].As used byremoved from a surfacerespectis occurring.condi-equalloads found to occur in the dry wear of steelsThe term "wear rate"most Tribologists,the ambientand rough whilstfor mild wear processesat removing materialnunof the surfacesrange of wear rates for each class of wear.the severe wear processeseffectiveagainstdeeply tornoftenwithexaminationmust be emphasized thatItto severe wear processeshowever,by reactionmicroscopicsmooth surfaces,is small,,10-4to the surfaceTypically,ments in which wear ratescan readilypresentingarea over which wear can occur than the othera much smallermember (which is normallycombinationsthe moving surfaces).of the same material,may be the same as for the stationaryfrom the whole of the wear track.much more difficultfavouredpairsof the moving surfacesurface,the removed volume comesMolgaardwillforthe wear ratetrackto measure by the normal distanceslidingone of the pairHence although,Quite clearly,by most wear researchers.such asymmetricalbe measured involvewear experi-[ll]wear ratestransducerhas pointedhave very differentaremethodsout thatheat flowsfrom

the scthan thatAny heatingof sliding).thatsuch as two annuli.preferencewearingslidingforwear.the pin-on-Providedin such geometries,situations(sinceitdue to differentitas the matingtime is a less fundamentaldistanceeffectsintoheat flowsto practicalremoved per unitHe suggestspairs,researchingis made for the differentas relevantThisgeometry has been found to be a difficultThe volume removed per unitannuli.involvesquantitythe speedspeeds can be allowedin the heat flow analysis.Having given an accountof Archardthatsymmetricalgeometry by most Tribologistsis probablyeach member.the wear.Hence, the continuedwhich to deal.allowanceforthishowever,intoand possiblytemperaturesshould be made withExperimentally,one withof heat at the interfaceand Hirst[9],most of Burwellthe simplerletand was probablyof each surfaceThis titlenamely one involvingmechanism of frictionare then transferredin a "back -transfer"a given pairBurwelloffcases ofTheirfirstclassimpliesaclearlyadhesion between the surfaces,'and Strangby adhesiveto the oppositeaway to become wear particlesare specialthe Bowden and Tabor [12] cold-weldingin mind.being pulledthe grounds for maintainingand severe wear.Wear".proposed withand severe wear classifications[8] classificationsof mildthe name "Adhesivemechanism of wear,know thatus considerand Strang'sclassificationswas assignedof the mildsurface,or they returnmechanism.of surfacesItforces.to theirpartsThese fragmentswhere they eitheris pertinenthas actually[8] envisageoriginalbreaksurfaceto ask, how does oneworn in thisway?

6Opticalmicroscopywhether transferexperimentsrevealrough surfacesor back-transferWarren [13],the earlyare difficultfor transferto perform.to followa transferwas prodigious.is clearclassificationthatitby Tabor [S],clean,(ii)step by step,in investigatingis not justifiedfrom the more readilyA he frictionaljunctionshow the interfaceadhesion itself)how the junctionitselfimportantin establishingthey willaffectso formed is primarilythe naturewillunder shear.a quantitativethe amount of surface(forcomeand the forceresponsibleof a pull-offsliding);These factorsandare clearlymodel for Adhesive Wear, sincematerialforof the atomic forcesdeforms under the actionforceconsidersthe atoms atforcesoccur at these pointsor a tangentialrupturesFor the present,under normal load,Tabor [S] discussesproposedas (i) Adhe-Tabor [5] mainlyand thus interatomicSome adhesion willforce.at the interface;in which,be in contactthe unificationof both.be concerned only with Adhesive Wear.(forhas shownthat wear can be classifiedand (iii)to break the interfacialforceto investigateWear" classificationat which to introducesince he maintainsNon-Adhesive,some pointsintobutof "Severe Wear".This is a good pointwe willAnalysiswear processis no differenttracerRecent work by Eiss andThe expense involvedthe mechanisms behind the "Adhesivebut notand back-transfer,stages of the wear of polychlorotrifluorethylene,but the effortsive,Radioactivewhere they used Neutron Activationto be possiblewhen ithave been formed,has occurred.have been used to testsuch experimentsitwillremoved duringsliding.

As yet,ratethere has appeared no theoryin terms of these factors,especiallysome promise,new techniquesand opy,and "Run-in"By thisdefinition,itby localis cleardefinition,as wellofbetween thenamely "scuffingis materialtheas the more conventional(such as is oftenalthoughhe does accept thatplasticflow at one asperity.considersthatabout some clarificationseen inconsiderstheboth Tabor [S] and Ludema [6]of the mechanisms and initiationto be "Severe Wear".there would be no contradictionWear", as used by BurwellweldingLudema [6] can now includefailuremust involveof what he (the author)[14] definitionLudema [6] does not see adhesion as beingfor scuffing,authorto anyhe seesflow whether or not thereof severe damage and wear lossof scuffingand theforms of wear in a wide spectrum ofa more generalof surfacetheirlookingmakes -no referenceschemes, althoughby plasticThe presentthatoxides,of surfacesthe only requirementhave broughtparticularlydamage characterizedinitiationwithinstruments!as two relatedMachine experiments).initiationanalysts,review of "Scuffing"He suggestsloss or transfer".microscopicyoung surfaceLudema [6] does not accept Dyson'sas "grossis a roughening4-Ballfilms,classificationother wear forms.slidingin his recentof surfaceof the previousof theirthe wearwould seem to be an approach withto scnne brightapplicationsLudema [6],"scuffing"but itsuch as Auger and X-Ray Photoelectronfor practicalfunctionof Adhesive Wear expressingand Strangas used by Dyson [14] and Ludema [6],in terminologyif[8] and Tabor [S],were consideredItis maintainedthe terms "Adhesiveand "Scuffing",to be slightly

forms of "Severe Wear", as defined by Archard and HirstdifferentThe differencescould possiblyand non-adhesive contact,lie in the relativeas being equivalentamounts of plasticdeformation which occurs between contactingassumes that "severe wear" involves plasticelasticamount of adhesivewhich could be interpretedto saying that the differencesand elasticlie in the relativedeformation,[9].surfaces.If onedeformation and "mild wear"then one can see how all the various forms of weardescribed by Burwell and Strang [S] and Tabor [S] as well as "Scuffing"and "Run-in",as proposed by Ludema [6], should in principal,in terms of relativedominance of eitherbe describablesevere or mild wear in any givensituation.Unfortunately,proportionsit is difficultto specify the effectof relativeof mild and severe wear upon the observable characteristicsof contact resistance,composition and size of wear debris,and thetopography of the worn surfaces (although this could be the subject offuture research!).all the classificationthe classifierAnother, and perhaps more important,systems other than Archard and Hirst'sknowing the mechanisms of wear [4].how the various classifications4.For the sake of brevity,can be reconciledA full[9] involvediscussion ofcan be found in referencean attempt is made in Table 1 to summarizethe main conclusions of that discussion,classificationproblem is thatand also include the recentsystems proposed by Tabor [S] and Ludema [6].

Table 1:--.Canparison of the Various Classificationsof Wear -Burwell 4Strang [8] --- AdhesiveWearQuinnCorrosiveWear. Mild Wear141Severe nMechanisms ofmild and severewear(See Section 1.4)FrettingCavitationIErosionMixtures ofAdhesive andNon-AdhesiveWearIt is possiblytionunfairas being equivalentto Ludema to include his "run-in"to corrosiveenvisages some surface fatiguerun-insurfaces,Nevertheless,and abrasion being involvedas well as chemical interactionit would seem that "Scuffing"severe wear and "run-in"interestingor mild wear alone.is essentiallyObviously,tion,over Quinn'sespeciallycontroversy[4] placingClearly,hein obtainingto form protectivefilms.is one form of adhesive orcorrosiveto note that Tabor sees no differenceCorrosive Wear.classifica-stillallor mild wear.It isbetween Adhesive andexistsover Wear Classifica-other types of wear under

10the category1.4.'Mechanisms of Wear", and thisIt does seem, however,be a simpler1.3thatclassificationitwillis generallythan thatthereshouldand Strang[8].The Laws of Weargeneral wear behaviorof surfacesThey should be compared withwere firstdiscoveredare proportionalthe apparentleadingarea of contact.unaltered.believein terms of plasticit willin these terms.to validateor otherIthypothesissevere wear.and topographysince the frictiondeformationThe proofthe similarityflatofremainis most striking,[15] of surfaces,the wear laws are also explainedthatin the experimentshypothesisof friction,i.e.,is only applicablesteelof frictiondoes not change,deformation[15] elasticsurface.undertakenoxide filmsthe plasticto surfacestheorybeing based on a model polymervery low speeds on a toolIn fact,laws have been "explained"were not present,of Archard'swhichof wear rateof the surfaces[12] and elasticof time beforeof frictionand the forceof the surfacesthat,contaminantsmost unsatisfactory,the natureis almost certainthe plasticsurfaceLaws of Friction,load and both are independentfrom this,deformationof themotion.in the 15th Century.Apartonly be a matterrelativeHowever, the proportionalitythe compositionone toduringsince both the wear rateto the normal appliedprovidedessentiallywhich interact[1950] descriptionthe well-establishedload is only true providedi.e.,recentby Leonardo da Vincithey are very similar,tionin Sectionagreed thatproposed by BurwellThe laws of wear are a comparativelywithbe discusseddeforma-exhibiting100%of frictionsurfaceslidingisatHence the "explanations"

11of frictionalsupporters1.4forceso far produced may be less universallyof these explanationsthan therealize.The Mechanisms of WearThe "explanations"in the lastsection,"mechanisms".involvedof frictionalmust clearlyIn other words,of wear.forcesduringboth frictiondeformationdominant factorto expect plasticin Severe Wear (in what follows,callWear" and "Scuffing""run-in"we willitSimilarly,is difficultin mild wear [as proposed by Tabor [S]].in more detail,what are the mechanisms of,withthatlubricants.Ifitinvolvesreactionthe oxygen in thatdue to the largeadditives,Wear [8] orfirstly,Let us consider,mild wear and,severe wear.Mild Wear clearlyappreciatedelasticto see how adhesioncan be involvedparticularto be a[8] and Ludema [6]respectively).Because of this,of friction.use thisshould be dominant in Mild Wear [or Corrosivewear [63.secondly,whereas elasticdeformationto stand for what But-well 4 Strangdeformationseems to be strongly[15] theoryclassification"Adhesiveto the "explanation"of friction,was the basic mechanism of Archard'stherefore,mentionedand wear must occur by relatedWe have seen how plasticis reasonable,sliding,have some relevancein the Bowden and Tabor [12] theorydeformationItvalidenvironment.the environment,Itamounts of airabsorbed and entrainedcontain"extreme-pressure"likely that the dominant reactioninis not alwaysmild wear can occur even under wet lubricatedthe lubricantsiswithwillin allconditionstypicalor "anti-wear"be between the metal

12(or itsoxide)involvesand the additive.reactionsMost typically,between the surfaceand any ambient oxygen, thatmild wear occurs through an Oxidationalhas evolved graduallyhowever, mild wearWear Mechanism.is,This mechanismover the past 20 years [16,17,18,19,20]and canbest be summarized as follows:In the initial(severe wear or scuffing)achieve a measure of conformity,of severalcomparativelystages,the surfacesso that the real area of contactlarge areas, each of which is about the size tobe expected from the Bowden and Tabor [12] plasticdeformationnamely, of the order of (W/p,), where W is the normal appliedis the BrinnellinstantHardness (expressed in unitsof contactto that proposed by Barber [Zl])If the eaucontactcontactingheatingwillregion.plateauof the surfaceessentiallyto thiswillcontactinguntilitis removed bythen the expansion of theFurthermore,the rate of oxidationnot be very differentto be theof thefrom that of the remaindertemperaturesunder these conditions.(i.e.,excessThis is,ofa mechanism for severe wear and we willaspect again in a laterplateau willIt then expands (in alarge for itdescribingGiven a sufficientAt any givennot be sufficientlytemperature)load and pmslow, or the loads are sois negligible,for the very low "hot-spot"over general surfacereturnspeed is comparativelytheory,so as to become a plateauand remain the only region of contactwear.course,of stress).one of these areas bears most of the load.manner similarconsistsparagraph.amount of frictionaloxidizeand the remainder of the surface.preferentiallyIt willheating,however, theto the other plateauxoxidize,during contact with

13the opposingthe generalat a temperaturesurface,surfacetemperature(Ts).has been proved by many investigators22, 231.They are extremelythe directionof sliding,heightsof about 2-3 m.dicularto the slidingfor wear throughthe oughwhich,and eventuallyThe contactinginteractionsheat and stresssimilarplateauxcriticalbecame the source of allcyclesasItexistingbecame unstable,the debrisfragmentsrough areas.plateauis the siteforsurfaces.allthe asperity/asperityAccordingoccurs by diffusionSince oxidationinwardsand (sometimes)plateauto grow in heightby*metalcontact.the increasesarea of the plateau.forions outwards,from the interfacemetal beneath each asperitytactreasonwear marks.heighttemperaturethatcrackon the opposingof a previouslyare the sitesfindperpen-are rough and strewnare no visibleitscracksto the fatiguewear mechanism, these asperitiesmuldtoone sees a possiblewere once partbetween two opposingTc.[18,parallelshow surfaceeach plateauupon reachingsee referencesan area of about 10e2nun2 andoftentherein excess ofwear tracksClearly,withsurroundingcontactin the non-contactingfinesomewhat similarcontactthese fragmentscracked,example,due to intermittentwould seem thatplateauof these plateauxwithmechanics.comes intoThe surfacesThe existencesmooth withThe plateauxfatiguewell(fortypicallysystems found in fracture(Tc) normallyin heightto the oxidationaloxidationat theof oxygen ionsone would expectthebetween the oxide and theIn the course of many passes,onewere spread over the whole con-When the plateaureaches a criticaloxide.

14film thicbess(c), the film becomes unstable and breaks up to formflakes and, eventually,[18,22] indicatesthe cellularrecently,wear debris.The electron microscope evidencea fatigue mechanism could be operating.sections discovered by Rigney, et al.Certainly,[24] and, morefound in some debris analyses by Allen [25], would seemto be consistent with the fatigue cracks and their regular spacingsat right angles to the directionThe oxidationalof sliding.wear mechanism, possibly to its o

1.1 Introduction Tribology is a subject which has suffered from a lack of precision over the terms used to identify its various constituents. Even the word "Tribology" itself has different connotations, according to the various authorities connected with the subject. The Jost Committee [1],