1、06FTM01The Effects of Super Finishingon Bending Fatigueby: G. Blake, Rolls-Royce - Transmissions and StructuresTECHNICAL PAPERAmerican Gear Manufacturers AssociationThe Effects of Super Finishing on Bending FatigueGregory Blake, Rolls-Royce - Transmissions and StructuresThe statements and opinions c
2、ontained herein are those of the author and should not be construed as anofficial action or opinion of the American Gear Manufacturers Association.AbstractSuperfinishingisatechnologythatpublicliteraturesuggeststohavepotentialforincreasedpowerdensity. Astudy was conducted characterize this technologi
3、es benefit to bending fatigue. Two sample groups werecreated, one from AMS6265, and the second from AMS6265 with super finishing.Bending fatigue was characterized using two test methods, Single Tooth Bending Fatigue (STF) and R-RMoore rotating beam. The STF specimen was designed such that the tooth
4、geometry replicated the normalcross section of a Rolls-Royce high power spiral bevel pinion gear. The design then allowed for the specialprocessing of the STF gears to be common to that of the spiral bevel gear.Twouniqueheatlotsofmaterialwereusedforeachsamplepopulation. Eachheatlotwasprocessedthroug
5、hmanufacturing as a separate machining batch. Thus, a minimum of two carburization and hardening loads,two shot peen batches and two super finishing cycles (if applicable) were processed per sample group.Sixty-twobendingfatiguetestpointswerecreated. DataanalysisofSTFandRRMooretestdataconcludedno sta
6、tistical difference between the bending fatigue strength the two populations.Copyright 2006American Gear Manufacturers Association500 Montgomery Street, Suite 350Alexandria, Virginia, 22314October, 2006ISBN: 1-55589-883-11The Effects of Super Finishing on Bending FatigueGregory Blake, Rolls-Royce -
7、Transmissions and StructuresIntroductionSuper finishing is a technology that public literaturesuggests has potential for increased power densityRef 1. A study was designed and conducted tocharacterizethistechnologysbenefittobendingfa-tigue. For the study, two sample groups werecreated: (i) AMS6265 a
8、nd (ii) AMS6265 with superfinish.Bending fatigue was characterized using two testmethods: Single Tooth Bending Fatigue (STF) andRR Moore rotating beam. The STF specimen wasdesigned so the tooth geometry replicated the nor-malcrosssectionofaRolls-Roycespiralbevelgeardesign. The design then allowed fo
9、r the specialprocessingoftheSTFgearstobecommontothoseof the Rolls-Royce spiral bevel gear design.Twouniqueheatlots ofmaterial wereusedforeachsample population. Each heat lot was processedthrough manufacturing as a separate machiningbatch. Thus, a minimum of two carburization andhardening loads, two
10、shot peen batches, and twosuper finishing cycles (if applicable) were pro-cessed per sample group.A detailed metallurgical evaluation of the speci-mens was performed to characterize the materialand compare it to that used in the Rolls-Roycespi-ral bevel gears.Data analysis of STF and RR Moore test d
11、ata con-cludednostatisticaldifferencebetweenthebendingfatigue strength of the two populations.NomenclatureCps Cycles per secondLCF Lowcyclefatigue,cyclestofailureless than or equal to 1e4 cyclesGRI Gear Research Institute, locatedwithin the Applied Research Lab-oratoryatThePennsylvaniaStateUniversit
12、yHCF High cycle fatigue, cycles to fail-ure greater than or equal to 1e4cyclesHL Raw material heat lotN Number of fatigue cyclesPowerdensityThe ratio between transmittedhorsepower and power trainweightR Stressrangeratio(fullreversingisdesignated as R = -1 and zero tomaximum is designated as R = 0)RR
13、 Moore A rotating beam test method thatproduces a full reversing stressrangeSF Super finishingS/N Serial number of specimen withina populationSTF Single tooth bending fatigue inwhich a special spur gear designis usedas afatiguespecimenanda pair of teeth is testedSuperfinishingAn isotropic chemical p
14、rocessperformed to achieve improvedsurface finishT Bulk temperature of specimenduring test _FExperimental ProceduresThis section contains the methods and proceduresusedtoplan,conduct,andverifythefatiguetesting.TheSTF gear design details along withtest riglimi-tations are also discussed in this secti
15、on.Number of STF Test Points and LoadLevelsThe number of test points and load levels were de-signed around the general relationship betweenbendingstress andlife. Thegeneral bendingstressversus life relationship appears in several literaturesources, and a summary is listed as follows (seeRefs 2-7):S
16、Life distribution is logarithmic.S Scatter is in the life direction between 1e4 and1e6 cycles.2S Scatter is in the stress direction beyond 1e6cycles.S Life increases as stress decreases and life de-creases as stress increases.S The slope of the stress versus life curvechanges slope and becomes almos
17、t horizontalbetween 1e6 and 1e7 cycles.S Infinitelifeisdefinedat1e7cyclesandtestsrun-ning declared as run outs beyond this point.S Lifecyclesgreaterthan1e4areconsideredtobehigh cycle fatigue (HCF).Sixtest pointsat threeloadlevelsinthefinitelifere-gime were created to characterize variation in lifebe
18、tween 1e4 and 1e6 cycles at each of the givenload levels. Fourteen or more test points werecreated tocharacterize thevariation instress intheinfiniteliferegime.Fatiguepointsateachlevelwerecreatedfrombothheatlotstoavoidbiasingoneloadlevel due to heat lot variation. Four tests were con-ductedfrom each
19、individual STFspecimengear.Tofurther avoid bias, all four possible tests from asingle gear were not conducted at the same loadlevel.Anillustrationofthetestloadlevelstocharac-terize the stress versus life relationship is shown inFigure 1.14 points, 2 heat lotsFigure 1. Illustration of test load level
20、s,number of test points, and expecteddistribution.Specimen Description and TestProcedureThedelimitingfactors,orboundaries,ofthebendingfatigue project include the material types and sur-face technologies evaluated. In addition, the basictesting methods were also delimiting factors. Allspecific factor
21、s are as follows:S Forged bar stock per AMS6265 specificationused for manufacture of specimensS Common carburize and hardening cycle usedfor all common specimen designsS All specimens shot peened per Rolls Roycelegacy specificationsS Fatigue test data generated using single toothbending and rotating
22、 beamS Tests performed at room temperature, approxi-mately 75_FS Fatigue tests suspended at 1e7 cycles and de-clared as runoutsS Super finishing process used for specimens aspreviously developed for the Rolls-Roycespiralbevel gearsS Baselineforsurfacetechnology(superfinishing)characterized is AMS626
23、5 ground and shotpeenS Full form ground STF specimens, root, root ra-dius, and flanksThe minimum number of STF gears required isequal to the number of fatigue test points requireddivided by the number of test sections per STFgears, as shown in Table 1. Additional STF gearswere necessary for metallur
24、gical evaluation.Table 1. Required number of STF gears forthe project.Endurance test pointsto be createdA 14Test points per loadlevel above endur-anceB 6Number of load levelsabove enduranceC 3Total number of teststo be createdT=A+(B*C) 32Tests per gear D 4Number of materials(populations) to betested
25、E 2Number of gears formaterial evaluationper materialF 1Minimum number ofgears required forprojectT*(E/D)+(E*F) 183The operating limits of the test rig were consideredin the specimen design. The STF gear face widthandnumber of teethweresizedbasedon themaxi-mum operating limits. Select teeth were red
26、uced inheight to allow access for the load and reaction an-vils during testing. Each STF specimen gearcreated a maximum of four test points.STF Test Setup and Fixture CalibrationThe STF testing was conducted at Gear ResearchInstitute located in the Applied Research Laborato-ry at PennStateUniversity
27、. Atest consistsof aloadbeingappliedontheleftsideofagear toothnumber1whileastatic anvilreacts theloadontherightsideof toothnumber4. Thus,twoteetharefatigueddur-ingthetest, as showninFigure2. The desiredanvilcontact point is in the upper 25% of the active pro-file.Figure 2. Layout illustration of STF
28、 test rigsetup.A rigorous test rig and fixture calibration was per-formedprior totesting. A gageconsistingof 10indi-vidual straingages was placedat themidfacewidthof one tooth root radius on the loaded side, asshown in Figure 3. The gear was then placed inthefixture with the instrumented tooth orien
29、ted up.Strain versus load was recorded for each of thestrain gages. The applied load was incrementedfromzerotoamaximum.Theprocesswasrepeateda minimum of three times to demonstrate repeat-ability of measured strain. The instrumented toothwas then oriented down and the strain versus loadsequencerepeat
30、ed.Thereactionanvilof thefixturewasthenshimmeduntiltheupperreadingsequaledthelowerreadings.Thesevalues werelater usedtovalidate the STF finite element model results notpresented in this paper.The maximum load frequency is a function of thetest rig and fixture, and specimen is determined foreach setu
31、p. A second STF gear was used todeter-mine the frequency for estimating the total testingduration. Figure 4 shows an acceptable load waveform at 20,000 lb and 15 Hz. The actual frequencyfor each test varied based on the maximum thatcould be achieved while maintaining a “smooth”load wave form.Figure
32、3. Strain gage placement.Figure 4. Load cell response at 15 Hz and20,000 lb.Thetemperatureof anSTF gearwas measuredus-ing a physical-contact temperature probe. Themeasurement was made on the end face, near theroot radius of the upper fatigued gear tooth. Thegear had completed 2.7e4 cycles at 20,000
33、lb loadat the time of measurement. The measured STFgear temperature was 74_F.Manufacturing ProcessingThe STF and RR Moore specimen manufacturingdetailsarediscussedinthissection.TheSTFgearswereprocessedinthesamemannerthat theRolls-Roycebevelgearswere.TheRRMoorespecimenswerealsoprocessedinasimilar man
34、neras thepro-4duction bevel gears. The applied case depth of theRR Moore specimens had to be reduced from thatof theSTF gears duetothecross section. Theheattreat, shot peen, and super finish cycles previouslydevelopedfortheRolls-Roycespiralbevelgearde-sign were used for the STF gears. STF gearmanufa
35、cturing was controlled and monitored tominimize variation. Specific STF gear items of con-trol and key features were as follows.S Two heat lots of material were used for eachsample group. Raw material details are listed intheMetallurgical Investigation sectionof this re-port.S Two machining lots per
36、 raw material heat lot re-sulted in two carburization and hardening loadsper sample group.S AMS6265STFgearswereheat-treatedusingaRolls-Royce production heat treat and cycle.S STF gears were shot peened using the sameprocess and vendor used for production bevelgears.S STF gears were super finished us
37、ing the sameprocess and vendor used for production bevelgears.S Stock removal was verified and recorded 100%for each tooth space.STF Gear Manufacturing ProcessA separate machining lot was created for each rawmaterial heat lot. A common basic manufacturingprocess for theSTF gears was definedas showni
38、nTable 2.Thedevelopedcarburizeandhardening cycleusedfortheproductionbevelgearswasusedfortheSTFgears, as shown in Table 3.The entire gear tooth surface and end faces wereshot peened using a dual cycle. The first of twocycles consisted of MI230 shot, 200% coverage,and an intensity of 0.013 to 0.018A.
39、The secondcycles consisted of MI110 shot, 200% coverage,and an intensity of 0.008 to 0.012A.RR Moore Manufacturing ProcessA separate machining lot was created for each rawmaterial heat lot. A common basic manufacturingprocess for the RR Moore specimens was definedas shown in Table 4. The same develo
40、ped shotpeen, and super finishing processes, used for theproductionbevelgearswereusedfortheRRMoorespecimens. The applied case depth of the RRMoorespecimenswasreducedfromthatoftheSTFgearsduetomaintainingthepropercasecoreratio.The finish case depth was 0.035 to 0.055 in. Therough and finish machining
41、was performed by Met-cut. Rolls-Royce carburized the AMS6265 speci-mens in the Plant 5 production heat-treat depart-ment. The same sources used to process the STFgears performed all other surface treatmentoperations.Table 2. STF gear basic manufacturingprocesses.AMS6265without SFAMS6265with SFLathe
42、LatheHob HobDeburr DeburrInspection InspectionCarburize CarburizeHarden HardenInspection InspectionLathe LatheID grind ID grindGear grind Gear grindInspection InspectionShot peen Shot peenSuper finishEDM EDMInspection InspectionTable 3. Developed carburize cycle.Operation Operation description10 Car
43、burize:0.070 in. to 0.074 in. cycle1700FFurnace cool to 1000FAir cool to ambient20 Harden:1500F, 0.85%C, 2 hrQuench in 110 to 190Foilfor10 min30 Temper:300F, 3 hr5Table 4. RR Moore specimen basicmanufacturing processes.AMS6265without SFAMS6265with SFRough machine Rough machineCopper plate Copper pla
44、teCarburize CarburizeHarden HardenFinish grind Finish grindShot peen Shot peenSuper finishSurface FinishSample roughness measurements were made onSTF gears and RR Moore specimens. Roughnessmeasurements were made on several STF gearsand RR Moore specimens with and without superfinishing. Roughness of
45、 the STF gears was mea-sured in the involute direction near the root radius,which was not directly accessible by the inspectionmachine. Roughness of the RR Moore specimenswas measured in the axial direction near the mini-mum cross section. All roughness measurementsshown in Table 5 were done post te
46、sting.Metallurgical InvestigationsMetallurgical investigation begins with the itemiza-tion of the raw material lab certification, carburiza-tionloads,andtheSTF gearsproducedfromthose.A summary of these data is listed in Table 6.ChemistryThechemicalcompositionofonesamplefromeachgroupwasdetermined. Sa
47、mpleswerecreatedfromthe partial teeth removed during manufacturing.Semi-quantitative XEDA measured values, in per-cent weight, along with the material requirements,arelistedinTable7.Table 5. Specimen roughness measurements.Material +iS/N Type of Requirements Measurements minprocessingptestRa Ra Rq R
48、v Rp RtAMS6265 w/o SF 1 STF 32.0 21.039 26.779 58.218 57.797 148.674AMS6265 w/SF 6 STF 4.0 4.094 5.503 13.379 10.364 33.631AMS6265 w/o SF 65B11 RR Moore 32.0 30.940 39.047 74.247 86.581 250.577AMS6265 w/SF 65A08 RR Moore 4.0 1.006 1.354 2.966 2.222 14.944Table 6. STF raw material heat lot and carbur
49、ization load data.Material Heat lot # Carb lot # Bar dia (in.) S/NsAMS6265 + SF #1 0087683-01-1 10 2, 3, 5, 6AMS6265 + SF #2 0090441-01-0 9.5 11, 12, 13, 31, 32 33AMS6265 #1 0087684-05-0 10 1, 2, 3, 4, 5AMS6265 #2 0087684-02-1 10 6, 7, 8, 12 17Table 7. Chemical composition.Si Mo Cr Mn Ni Cu FeW/o SF 0.3 0.0 1.4 0.8 3.2 0.2 94.1W/ SF 0.4 0.1 1.3 0.7 3.2 0.2 94.19310 spec 0.15 - 0.35 0.08 - 0.15 1.0 - 1.4 0.40 - 0.70 3.0 - 3.5 0.35 BAL6Hardness Traverse, Case Depth, andMicrostructureThehardnesstraversemeasuresthedepthofcase.Thedepthof caseis defined
