AGMA 05FTM15-2005 Repair of Helicopter Gears《直升机齿轮的维修》.pdf

1、05FTM15Repair of Helicopter Gearsby: S. Rao and D. McPherson, The Pennsylvania State Universityand G. Sroka, REM Chemicals, Inc.TECHNICAL PAPERAmerican Gear Manufacturers AssociationRepair of Helicopter GearsSuren Rao and Doug McPherson, The Pennsylvania State Universityand Gary Sroka, REM Chemicals

2、, 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.AbstractIn order to reduce costs by extending the operational life of the sun and input pinion gears of a helicoptertr

3、ansmission, scraped gears were subject to a superfinishing process. This process was found to removeminor foreign object damage by uniformly removing a minimal amount of material on the gear teeth, whilemeeting original manufacturing specifications for geometry. The process also resulted in enhanced

4、 surfacequality and did not exhibit detrimental metallurgical effects on the surface or sub-surface of the teeth. Theprocess was also found to eliminate gray staining, an early precursor to pitting. This paper describes theresults of the helicopter gear repair project and includes the geometry and m

5、etallurgical evaluations on therepaired gear. Further effort to characterize the durability and strength characteristics of the repaired gear isongoing.Copyright 2005American Gear Manufacturers Association500 Montgomery Street, Suite 350Alexandria, Virginia, 22314October, 2005ISBN: 1-55589-863-71Rep

6、air of Helicopter GearsSuren Rao and Doug McPherson, The Pennsylvania State Universityand Gary Sroka, REM Chemicals, Inc.BackgroundAt each depot level maintenance interval (900 hoursflying time for the Input Pinion and 1800 hours flyingtime for the Sun Gear) the gears are thoroughlyvisually inspecte

7、d. This inspection is carried outwith up to 10x magnification, and gears are either,scrapped, returned to service, or subjected to re-mediation based on the inspection findings.Figure 1 shows typical minor FOD. Remediation isaccomplished by hand polishing with 400 grit paperin accordance with approv

8、ed repair specifications.The intent of this process is that no (i.e., so little as tobe negligible) material is removed, and that thegears are not “reworked.” Figure 2 shows typicalgray staining and early stages of pitting. Thesegears are routinely reassembled into the transmis-sion with potential p

9、remature pitting failure develop-ing as a result.Figure 1. Sun Gear with minor FOD, areacircledinredCH-46 Aft transmission Sun Gears and InputPinion gears are being scrapped at a rate of over$1.2 million per year. The Sun Gears are by far thehigher dollar value item, and most of this effort is di-re

10、cted at evaluating superfinishing as a method tosalvage some of these gears and reintroduce theminto the transmission.Figure 2. Input Pinion with early pitting (graystaining) on drive flanks below pitch lineReview of Superfinishing ProcessesThere are several commercial methods currentlyavailable to

11、superfinish the surface of parts that re-sults in a non-directional, isotropic, surface finish.These remove small amounts of material uniformlyover the surface of the parts, and can eliminatesome minor surface defects. Typically, these pro-cesses are used to impart a mirror, or near-mirrorfinish to

12、machined parts. It is proposed to use one ofthese methods to uniformly process gears that aredeemed suitable for remediation without introduc-ing non-uniformities in surface topography whiledoing so. Three of the commercially available massfinishing processes were evaluated as a part of thiseffort.

13、There were: MFI (Mass Finishing, Inc.),Extrude Hone, and REM Chemicals, Inc.A summary of the main characteristics of these pro-cesses is tabulated in Table 1.2Table 1. Evaluation of Superfinishing ProcessesProcess REM Chemicals, Inc. Mass Finishing, Inc. Extrude HoneHydrogen embrittlement None See N

14、ote 1 N/A (not an acid pro-cess)N/A (not an acid pro-cess)Selective etching atgrain boundariesNone See Note 2 N/A (not an acid pro-cess)N/A (not an acid pro-cess)Process effect on resid-ual stressNone See Note 3 No data available Adds slight compressiveresidual stressMaterial removal Depends on surf

15、ace fin-ish before processing.For aerospace qualitygears, stock removal isapproximately 0.0001”per surface.Varies. For aerospacequality gears very, verylittle material is removed(no quantitative valueavailable).Can be controlled tonearest 0.0001”.Finish achieved Bright finish (about 2 mi-cro-inch Ra

16、). Can alsocontrol process to re-move less material andtarget any required fin-ish.Bright finish (can go to0.02 micro-inch Ra).Can also control pro-cess and target any re-quired finish.Depends on finish be-fore processing. Can re-duce Raby 8:1. Limitabout 4-6 micro-inch RaEffect on temperature ofrun

17、ning gearsReduces temperature inautomotive & aerospacetransmissions.No data available. No data available.Process time (qualita-tive)Rapid Slow SlowExperience in superfi-nishing gearsSignificant Minimal MinimalProcesses are applicable to the following materials:Case hardened Yes Yes YesThrough-harden

18、ed Yes Yes YesNon-hardened Yes Yes YesNon-ferrous Yes Yes YesPlated Yes. Removes plating. Yes. Can remove or bur-nish plating.Yes. Can remove or pol-ish plating (if plating isthick and uniformenough).Black oxide coated Yes. Removes black ox-ide coating.Yes. Can remove or bur-nish black oxide coating

19、.Yes. Can remove or pol-ish black oxide coating(if coating is thick anduniform enough).Notes:1. REM Chemicals, Inc. has conducted studies that verify the mild acids (pH = approximately 5.5) used in theirprocess, coupled with the other details of the process, do not cause hydrogen embrittlement.2. RE

20、M Chemicals, Inc. has conducted studies that show their process does not selectively attack grainboundaries.3. REM Chemicals, Inc. has measured residual stress on parts before and after processing and verified theirprocess has no measurable effect on residual stress.(Results of the studies noted abo

21、ve can be obtained from REM Chemicals, Inc.)The MFI process is a vibratory honing process. Thehardness and texture of the abrasive media can becontrolled to achieve any desired surface finish,down to a bright mirror finish (0.02 micro-inchesRa). With proper media selection, thin coatings canbe polis

22、hed without being removed. This is a timeconsuming operation. The Extrude Hone processinvolves honing the surface of the part with putty likemedia charged with abrasive particles. This is a cut-ting process rather than a burnishing/honing pro-cess. The surface finish that can be achieved is afunctio

23、n of the surface finish before processing with3the functional limit for finest finish is approximatelyin the 4 to 6 micro-inches Rarange. This methodrequires tooling specific to each part to ensure uni-form stock removal. The REM Chemicals processis a chemically accelerated vibratory finishingtechno

24、logy. A mildly acidic phosphate solution pro-duces a soft conversion coating, which is easily re-moved by the rubbing of the non-abrasive media.he exact blend of chemicals and the size and shapeof the media are selected on the basis of the materi-al being processed, the size and shape of the part,an

25、d the desired final surface finish.The MFI process requires long process time to pro-duce fine finishes. The Extrude Hone process re-quires special tooling to ensure uniform stock re-moval, and can produce only a moderately finefinish. Thus, of the three processes evaluated, theREM Chemicals process

26、 offers the best advantagein producing a fine finish (1-2 micro-inches Rarange), rapid processing time, and controlled, mini-mal, stock removal. Based on this analysis (a de-tailed analysis was not considered a part of the ef-fort) and considerable experience with the REMChemicals process for superf

27、inishing gears(1,2,3),itwas selected for this effort.Pre- Finishing EvaluationsTwo Sun Gears and two Input Pinions that had beenscrapped due to major FOD were provided for ex-amination. These gears also exhibited minor FOD(edge not detectable with the point of a sharp probe,a typical example is show

28、n in Figure 1) and early pit-ting (gray staining, a typical example is shown inFigure 2) was detected on the Input Pinions, makingthese gears good candidates to investigate repairprocedures.Metallurgical Characterization Pre-SuperfinishingA metallurgical section was cut from one of the sungears thro

29、ugh an area with minor FOD damage.This section was polished, and examined for evi-dence of sub-surface cracking resulting from thedislocation when the foreign object passed throughthe mesh. A high magnification photomicrograph ofthis section is shown in Figure 3. No evidence ofsub-surface cracking c

30、ould be detected. Thus,damage of this limited magnitude is a satisfactorycandidate for remediation by superfinishing the sur-face.Figure 3. High magnificationphotomicrograph of section through area onSun Gear with minor FOD. No sub-surfacecracking associated with the damage is evi-dent.A metallurgic

31、al section was cut from one of the InputPinions through an area with early pitting (graystaining). A high magnification photomicrograph ofthissectionisshowninFigure4.Figure 4. High magnification photomicro-graph of section through area on Input Pinionwith early pitting (gray staining). Damage isconf

32、ined mainly to the surface with only mi-nor (less than 0.0001 inch) penetration intothe substrate.Examination of the polished section showed that theearly pitting (gray staining) was confined mainly tothe black oxide coating and penetrated less than40.0001 inch into the substrate. Again, damage ofth

33、is limited magnitude is a satisfactory candidate forremediation by superfinishing the surface. TheREM Chemicals process will remove the coating,and can be controlled to remove the top 0.0001 inchof the substrate.Dimensional Characterization Pre-SuperfinishingA comprehensive dimensional characterizat

34、ion wasconducted on one Sun Gear and one Input Pinion.Standard dimensional characterization of gears in-volves checking profile, spacing, index, run out, andthickness variation at the center of the face width.To get a more complete picture of the effect of REMChemicals superfinishing on the gears fo

35、r this effort,these items were measured at the center of the facewidth and 1/4 inch from the ends of the face for eachgear. On the Input Pinion, these measurementswere conducted on each of the right hand and lefthand helices. The results of the dimensional char-acterization prior to superfinishing a

36、nd after super-finishing is presented comparatively in Table 2 in thefollowing text.Surface Roughness CharacterizationPre-SuperfinishingSurface roughness measurements were conductedon one of the sample Sun Gears, prior to superfin-ishing. A surface roughness of 7 micro-inches Raand 18 micro-inches R

37、awas measured along thelead direction and along the profile direction,respectively.Based on the fact that the damage exhibited on boththe Sun Gear and Input Pinion was satisfactory forrepair by superfinishing the Sun Gear, the Input Pin-ion, and a segment from the second Sun Gear,which had been sect

38、ioned for metallurgical ex-amination, were sent to REM Chemicals, Inc. for su-perfinishing. The segment of the sectioned gearwas sent to provide a sample for metallurgical char-acterization and residual stress measurement aftersuperfinishing. REM Chemicals, Inc. masked allareas on the gears except t

39、he teeth. Processingwas carried out using media and chemicals thatREM Chemicals, Inc. has determined are suitablefor processing AISI 9310 steel gears of the generalsize and configuration of the gears for this effort.Table 2: Dimensional Changes Due to Super FinishingParameterSun GearP/N 107D2256-7Me

40、etSpecInput PinionP/N A02D2059-MeetSpecTooth Thickness Reduced 0.00014 Yes Reduced 0.0003 (1)Lead Added crown and taper totalvariation less than 0.00005per flank.Yes None Measurable YesProfile Increased Tip Relief 0.0001 Yes Increased Tip Relief 0.0001 (2)Index Variation None Measurable Yes None Mea

41、surable YesPitch Line Runout None Measurable Yes None Measurable YesTooth Spacing Varia-tionNone Measurable Yes None Measurable (3)Tooth Thickness Vari-ationNone Measurable Yes None Measurable YesProfile Hollow None Measurable Yes Broke the edges of areas withreverse curvature, reducingthe maximum t

42、o less than0.000075 inch per degree roll.(4)Notes:(1) Tooth thickness of Input Pinion was below minimum OEM specification before super finishing.(2) Profile on drive flank of Input Pinion did not meet OEM specification due to excess tip relief beforesuper finish. Profile on coast flank met OEM speci

43、fication before and after super finish.(3) Maximum tooth spacing variation on the Input Pinion exceeded OEM specification before super finish-ing. Super finishing was not able to correct the tooth spacing error.(4) There were select areas on the Input Pinion with profile hollow (reverse curvature) i

44、n excess of OEMspecification before super finish. Super finishing generally corrected this.5Superfinishing Process DetailsREM Chemicals process is produced in vibratoryfinishing bowls or tubs. A proprietary active chemis-try is used in the vibratory machine in conjunctionwith high-density, non-abras

45、ive ceramic media.When introduced into the machine, this activechemistry produces a stable, soft conversion coat-ing on the surface of the metal gears being pro-cessed. The rubbing motion across the gears de-veloped by the machine and media effectively wipesthe conversion coating off the asperity pe

46、aks of thegears surfaces, but leaves the valleys untouched.(No finishing occurs where media is unable to con-tact or rub.) The conversion coating is continuallyre-formed and rubbed off during this stage produc-ing a surface smoothing mechanism. This processis continued in the vibratory machine until

47、 the sur-faces of the gears are free of asperities. At thispoint, the active chemistry is rinsed from the ma-chine with a neutral soap. The conversion coating iscompletely rubbed off the gears to produce the iso-tropic superfinished surface. In this final step, com-monly referred to as burnishing, n

48、o metal is re-moved.The gears used in this study were superfinished in a10 cubic foot vibratory bowl using non-abrasivehigh density ceramic media (3 mm x 6 mm angle cutcylinders) and a proprietary active chemistry. A wit-ness coupon was processed simultaneously alongwith the gears. After 1.25 hours,

49、 0.00010 inch ofstock was removed from the witness coupon. Thegears were then burnished for 2.0 additional hours.After burnishing the gears were immersed in a rustpreventive, dried and returned for evaluation.Post- Superfinishing EvaluationsMetallurgical CharacterizationPost-SuperfinishingA metallurgical characterization was conducted ona section cut from the segment of a Sun Gear thatwas superfinished. The hardness profile obtainedin this characterization is shown in Figure 5. Sur-face hardness, depth to 50 HRC, and core hardnessmeet OEM Specifi