1、06FTM02Isotropic Superfinishing of S-76C+Main Transmission Gearsby: B. Hansen, Sikorsky Aircraft Corporation, M. Salernoand L. Winkelmann, REM Chemicals, Inc.TECHNICAL PAPERAmerican Gear Manufacturers AssociationIsotropic Superfinishing of S-76C+ MainTransmission GearsBruce Hansen, Sikorsky Aircraft
2、 Corporation, Mike Salerno and LaneWinkelmann, REM Chemicals, Inc.The statements and opinions contained herein are those of the author and should not be construed as anofficial action or opinion of the American Gear Manufacturers Association.AbstractIsotropic Superfinishing is a chemically accelerat
3、ed vibratory finishing process that is capable of generatingsurfacefinisheswithanArithmeticMeanRoughness(Ra)3min.Thisprocesswasappliedtothethirdstagespur bull gear and mating pinions along with the second stage bevel gears of a Sikorsky S-76C+ maingearbox. ThegearboxcompletedthestandardAcceptanceTes
4、tProcedure(ATP)anda200-hourendurancetest. Duringthesetestsnoise,vibration,andoperatingtemperatureswereshowntobesignificantlyreduceddue to the lower friction. This technology has since been flight certified and integrated into the S-76C+ withseveralaircraftincommercialservice. Adescriptionofthetests,
5、performancedataandageneraldescriptionof the process will be presented.Copyright 2006American Gear Manufacturers Association500 Montgomery Street, Suite 350Alexandria, Virginia, 22314October, 2006ISBN: 1-55589-884-X1Isotropic Superfinishing of S-76C+ Main Transmission GearsBruce Hansen, Sikorsky Airc
6、raft CorporationMike Salerno and Lane Winkelmann, REM Chemicals, Inc.IntroductionNoise and vibration control is a primary concern forthe transmission design engineer, and particularlyso in the design of a helicopter main transmission.Excessive vibrations generated by transmissionstypicallyresultinun
7、desirablenoiselevelsinhelicop-ter cockpits and/or cabins, which cause operatorand/or passenger aural discomfort and/or damageto sensitive on-board instrumentation. Cabin and/or cockpit noise/vibration abatement is of particularconcern in helicopters wherein the final stage of re-ductiongearingofthem
8、aintransmissioncomprisesone or more bull pinions interacting with a centralbull gear.For example, Sikorsky helicopters of the S-76 se-ries, e.g., S-76A, S-76B, S-76C, have a maintransmissionthatincludesthreestagesofreductiongearing:afirststageforeachengineoutputconsist-ingofhelicalgearing,anintermed
9、iatestageconsist-ing of spiral bevel gearing, and a final reductionstage comprising a central bull gear that inter-mesheswithrightandlefthandbullpinions(tocom-bine the inputs of the two engines that provide themotive power for the helicopter). Research hasshown that the cockpit and/or cabin noise le
10、vels ofS-76 helicopters are primarily the result of vibra-tions originating in the main transmission.Narrow band Fast Fourier Transform (FFT) analy-ses,A-weightedoctavelevels,andoverallDBAlev-elsrecordedinthecockpitsand/orcabinsofS-76A,S-76B, andS-76C helicoptersindicate thatinteriornoise levels are
11、 predominately the result of vibra-tions occurring atthe bullgearing meshingfrequen-cy of 778 Hz, as illustrated in Figure 1. The vibra-tions produced by the first and second reductionstages of S-76 main transmission gearboxes, i.e.,the noise levels generated by the helical and spiralbevel gearing a
12、s illustrated in Figure 1, occur athigher frequencies and typically are not significantrelative to the dominant noise levels produced bythe fundamental and first few harmonics of the bullgearing meshing vibrations.The gearbox vibrations resulting from bull gearmeshing are transmitted to the helicopt
13、er airframevia the transmission housing. The resultant air-frame vibrations generate noise in the helicoptercockpit and/or cabin. Abatement of such noise byacoustictreatmentofthecockpitand/orcabininteri-or is generally inefficient, and therefore, effectivenoise control solutions must be implemented
14、at thenoise source, i.e., the main transmission.To effectively abate such noise, it is necessary toidentify the primary causal factor(s) of bull gearingvibrations. The vibrations generatedby thehelicop-ter main transmissionmay beaggravated bymesh-ingbetweenmisalignedbullgearing,i.e.,thecentralbull g
15、ear and bull pinion(s). Previous efforts to re-duce the noise levels generated by intermeshingbetween misaligned bullpinions andthecentralbullgear included modifications to provide effectivebullgeartiprelief.Whilesuchmodificationsresultedinamodest reduction in bull gearing vibrations, the re-sultant
16、 reduced interior noise levels of S-76 heli-copterswerejudgedtostillpresentanunacceptablelevel of aural discomfort.1As a result, several avenues wereexploredtomorefully identify the design and operating parametersthatcausenoiseinducedbythe intermeshinggearsofadrivetrain.Onesuchavenuewastheinvestiga-
17、tionofSuperfinishingusingChemicallyAcceleratedVibratory Finishing. This process generates a verysmoothsurfaceontheflanksofgearsandhasprov-en its ability to reduce the coefficient of friction2,andinlaboratorytesting,thefrictionalcomponentofnoise known as Vibro-Acoustic Noise.32Figure 1: Chart showing
18、 that the predominant cabin noise is produced by the gearing meshingvibrations of the first few harmonics of the bull gear. Note that the highest cabin noise level ispredominately the result of vibrations occurring at the first harmonic bull gearing meshing fre-quency of 778 Hz.Vibro-acoustic noiseT
19、he gear meshing cycle consists of a largeamountof sliding between the teeth making contact. At thepitch point, the gears experience pure rolling, how-ever,theareasbeforeandafterthepitchpointexpe-rience much more sliding motion. At this interface,the teeth experience high contact stresses whichcan re
20、sult in low oil film thickness. When this hap-pens, the result can be mixed or boundary lubrica-tion conditions which lead to an increased coeffi-cient of friction. At this point, the coefficient offrictionis controlled either by thechoice of lubricantor thesurface finishof theteeth. This slidingand
21、/orshearing of intermeshing teeth has been theorizedas a significant contributor to the total noise signa-ture of a gearbox and in the last few years, thistheory has been proven in laboratory testsNO TAG.As such, it will be appreciated that by Superfinish-ing, the friction coefficient on the gear te
22、eth will bedramatically reduced, and so too, is the componentof noise produced by the sliding action of the inter-meshing gear teeth.Superfinishing using chemicallyaccelerated vibratory finishingChemically accelerated vibratory finishing incorpo-ratinghighdensitynon-abrasiveceramicmediaen-hances the
23、 performance of components that aresubjected to metal-to-metal contact or bending fa-tigue. The isotropic surface generated is uniquewhen compared to even the finest honing and lap-ping in that it has no directionality and is capable ofproducing a final surface roughness of 4 min. (0.1mm) Ra. It has
24、 a texture which consists of only ran-domscratchesandshallowdents. Whentheresult-ant surface has an Raof approximately4.0 min.(0.1mm) orlessandanon-directionalsurfacepattern,itis referred to as an Isotropic Superfinish (ISF),however, for this document Superfinishwill beusedto denote this process and
25、 surface topography.Figure 2shows anSEM imageat 1000Xof atypical3ground surface with an Raof approximately 10 min.(0.25mm) (top image) and a Superfinished surfacewithanRa2min.(0.05mm) (bottomimage) forref-erence. Note the directional asperities on thegroundsurfaceandthelackof directionalasperitieson
26、 the Superfinished surface. Only light scratchesand tiny dents are visible amongst smooth, pla-teaued areas of the Superfinished surface. It istheorized that these tiny dents facilitate lubricantretention during operation.4Such surfaces areunique in their remarkable ability to reduce fric-tion2, wea
27、r5, operating temperature6, andnoise3, as well as contact7and dynamic fatigue8when compared to similar surface finishes pro-duced by other techniques. There has been in-creasinginterestovertheyears inapplyingthispro-cess to aerospace gears since the industry is beingdriven to produce higher cycle li
28、fe gears at in-creasedpowerdensitieswhilestillallowingforlargesafety margins.Figure 2: Scanning Electron Micrograph (SEM) at 1000X of a typical ground surface with an Raapproximately 10 min. (0.25 mm) (top) and a Superfinished surface with an Ra 2 min. (0.05 mm)(bottom). Note the isotropic (non-dire
29、ctional) surface texture on the Superfinished surface ver-sus the directional asperities on the ground surface.4The Superfinish process is discussed in detail else-where7, but the following is a brief summary of theprocess. Superfinishisproducedinvibratoryfinish-ing bowls or tubs. A proprietary acti
30、ve chemistry isused in the vibratory machine in conjunction withhigh-density, non-abrasive ceramic media. Whenintroduced into the machine, this active chemistryproduces a stable, soft conversion coating on thesurfaceofthegear(s)beingprocessed.Therubbingmotion across the gear(s) developed by the ma-c
31、hine and media effectively wipes the conversioncoating off the “peaks” of the gears surfaces, butleavesthe“valleys”untouched.(Nofinishingoccurswhere media is unable to contact or rub.) The con-versioncoatingiscontinuallyre-formedandrubbedoff during this stage producing a surface smoothingmechanism.
32、This process is continued in the vibra-tory machine until the surfaces of the gear(s) arefree of asperities. At this point, the activechemistryis rinsed from the machine with a neutral soap. Theconversion coating is rubbed off the gear(s) one fi-nal time to produce the Superfinished surface. Inthis
33、final step, commonly referred to as burnishing,no metal is removed.The Superfinish process is a mass finishing opera-tion whereby tens or hundreds of gears can be si-multaneously processedinthesamemachine. If alloftherawgearsplacedinthevibratory machineareidentical at thestart, thenthey areall ident
34、ically fin-ishedattheendoftheprocess.Everygeartoothwillhave the same surface finish and geometry sincethe parts continually and randomly move throughthe vibratory machine and statistically experiencethe same chemical and media exposure. If onetooth on a gear has its tooth thickness reduced by0.0003
35、inches (0.0076 mm), then every tooth onthatgearandeverygearinthefinishingmachinewillhave its tooth thickness reduced by the sameamount. Therefore, there is no need for costly finalinspection of each and every gear as must be doneafter grinding or honing.Inaddition, thesimplicity oftheSuperfinishproc
36、essyieldsaveryrobustmanufacturingmethod.Vibrato-ry machines are run for years without any mainte-nance except for minor lubrication. The media isnon-abrasivesoitretainsitsshapeandsizeforlongperiodsoftime.Theimportantparametersthatcon-trolthesurfacefinishingoperationarethenumberofgears in the finishi
37、ng machine, the concentration ofthe active chemistry, theflow rateof activechemis-try,andtheprocessingtime.Alloftheseparametersareeasily controlled. Theprocess lendsitself toau-tomation,andhasbeencommerciallyusedoverthepast 19 years.S-76C+ Main Transmission TestingOrigin of the test gearsSikorsky Ai
38、rcraft Corporation supplied off the shelfS-76C+ main transmission gears to REM Chemi-cals, Inc. for Superfinishing. This process was ap-pliedtothethirdstagespurbullgearandmatingpin-ions along with the second stage bevel gears. SeeFigure 3 for a schematic of the S-76C+ drive trainand the gears proces
39、sed. These gears weremanufactured from AMS 6308 steel, carburized toexistingSikorskyAircraftCorporationspecificationsand precision ground. The goal was to Superfinishthese production gears to a Ra4min. (0.1 mm)across the active profile area of the teeth.Starting condition of the test gearsThefive ge
40、arswere acceptedproduction gearsandwere in typical aerospace precision ground condi-tion. The surface roughness of the gears wasmeasured by REM Chemicals, Inc. upon receiptand prior to processing using a Hommel T1000profilometerfittedwitha5micronradiusstylusandfollowing the ISO 4288 specification fo
41、r themeasurement of surface roughness. For easycomparison to the finished surface, a TracingLength (Lt) of 1.5 mm and a Wavelength Cut-off(Lc) of 0.25 mm was used in the measurement oftheas-receivedgears sincethisis thespecificationfor a non-periodic surface pattern and Rabetween0.8min(0.02mm) and 4
42、.0 min(0.1mm). Thesurfaceroughness results of the as-received, or startingcondition, gears prior to Superfinishing is listed inTable 1. A typical surface roughness profile anddata printout for an as-received gear is shown inFigure 4. This printout is from a third stage spurpinion gear.5Figure 3: Sch
43、ematic of an S-76C+ drivetrain which highlights the gears Superfinished during theproject.Table 1: Surface roughness measurements of the S-76C+ main transmission gears as-receivedprior to Superfinishing.Measurement pa-rameterSecond stage bevelpinions and gears (min.)Third stagespur pinion (min.)Thir
44、d stage bull gear(min.)Ra13 to 18 16 to 17 13 to 17Rz80 to 115 97 to 101 83 to 113Rmax109 to 167 119to129 133 to 141* AllmeasurementstakenatLt=1.5mmandLc=0.25mmusinga5micronradiusstylustip.Figure 4: A typical surface roughness measurement for an as-received (un-Superfinished)S-76C+ main transmission
45、 gear. The trace was taken along the profile and within the active pro-file region of the tooth flank on a third stage spur pinion. This measurement is typical of all as-received gears used in the project. The scale to the right of the profile indicates that each verti-cal segment represents 5 mm an
46、d each horizontal segment represents 100 mm.6Superfinish processTheSuperfinishingprocess consistedof thefollow-ing steps:1. Critical areas of the five gears such as bearingjournals, threads and bolt holes were masked inordertopreventstockremovalduringtheSuper-finish process.2. Theexposedsectionsofth
47、egearswerecleanedof grease, oil and contaminants in order to en-sure a uniform reaction with the active chemis-try.3. Superfinish (stock removal phase) using com-mercially available active chemistry.4. Burnish(cleaning) usingcommercially availableburnishing compound.5. Unmask critical areas.6. Rust
48、prevent using commercially available rustpreventative.Final condition of the test gears after Superfin-ishingAfterprocessing,thegearswerephotographedandfinal surface roughness measurements were takenofeachgear. Visualinspectionofthegearsshowedthat the directional asperities had been removedleaving a
49、 smooth and flat contact surface in itsplace. The appearance of the gears was brightandhighly reflective, indicative of a highly polished sur-face. Figure 5 shows a section of teeth from thethird stage bull gear after Superfinishing. Note thereflections of the other teeth across each flank.Figure 5: An image of a section of teeth fromthe third stage bull gear after Superfinishing.Note the reflections of the other teeth acrosseach flank.Thefinalsurfaceroughnessofthegearswereagainmeasured by REM Chemicals, Inc. after Superfin
