1、 92FTM12Noise Reduction in a Plastic and PowderMetal Gear Set Through Control of“Mean Involute Slope“by: Robert E. Smith, R.E. Smith _COMMUTATOR SEGMENTS 22nd show up in both of the tests mentioned _44th 2nd Harmonic above.2 3.1. Transmission Error Tests.Transmission error testing is usuallydone wit
2、h high resolution optical _ ilencoders and instrumentation that measuresmall deviations in the smoothness ofrotational motion. Typically, it is done R*,Gesae dRY STATUS,P*USZ_at relatively light loads. In this case *,STORZ_ _Ma_3eF 10tests were run at light loads and at vr,#_ _ _ m _ _-discernable d
3、ifference in the tooth totooth transmission error at either loadcondition. This showed that tooth I m :deflection under load was not a problem aR wand that tip relief on the teeth was not _vnecessary. Figure 3 shows the results ofa transmission error test. The firsthalf of Figure 3 shows the totaltr
4、ansmission error curve for threerevolutions of the 48 tooth _ear. One .wHt “can see a sine wave from each revolution x,STARTS|3SeHzz Yt_M!aS.3 ff_51)ta_rl_)results of the 48 tooth plastic gear.Where:It should be noted again that the slopeof each involute trace varies as the CA = actual pressure angl
5、eresult of runout in the part. This (degrees)slope, between the EAP and SAP, is noted = nominal (or set-up) pressurefor each trace (fBa). Another important angle (degrees)number on these charts is the value for MIS = Mean Involute Slope(fHam), the MIS. (These symbols are from dB = nominal (or set-up
6、) baseISO. There are no equivalent AGMA circle diametersymbols), ez,el = upper, lower roll angles overwhich the MIS is measured _4O 3.3. Experimental Results. In order toprove that MIS was a good measure of) quality gears, some pinions were reworkedto match the minus .0005“ tip MIS of the48 tooth pl
7、astic gears. The pinions weremade to have a minus .O00S“ to minus.0006“ tip MIS also.5.24“ The involute results are shown in Figure7. They are now basically the same .pressure angle as the 48 tooth gears.RESULTS- blO,O00 INCHLEFTFLANK RIGHTFLANK Figure 8 shows the improved results on4 3 2 i i 2 3 4
8、the transmission error test. Figure 9fEa 9.4 9.5 1.7 1.8 3 -0.? 5.5 0.2 shows the improved noise test. It isdifficult to see peaks in the spectrumthem (MIs) s.6 4.s that relate to the gear mesh and harmonicfrequencies and the subjective whine wasFigure S. INVOLUTE TESTS OF I2 TOOTH PINION gone.FROM
9、NOISY GEAR SETRESILTS - 1/10,000 IlLfli LIFT FLANK IIIIT FLANK4 3 2 I 2 3 4V_ tNa -4 -4 -9 -9.5 -9 -6 -1 -6e “ - - 15.8 fHom (MIS) -6.6 -5.5Picture 7. INVOLUTE TESTS 01: MODIFIEDRESULTS - 1/0,000 INCH 12 TOOTH PINIONLEFT FLANK RIORT FLANK4 3 2 1 _ 2 3 4 4000 .004fNa -6 -7.3 -7.5 -6.6 -4.8 -6.? -5.6
10、-4.8 , J_ I EEV _J4_, TOOTE GEARfHam (MIS) -6.0 -5.5 2000 .002 I _ i F_sure 6. INVOLUTETESTSOF 43 TOOTEGEAR _ /*_A /e_/_ /h_FEOM NOISY GEAR SET 0 0-2000 -,002-4OOO . .004l_flu .mmTlmIp_.mlmwm _ t_For the pinion: aT_ Tg_m4_ss|cm z_oe = 20MIS = +.0005“ moo .oo_dB = . 4698“ez = 40.000 at .573“ diameter
11、 I-“J _ ! TOOTH MESH_i = 5.24 at .4718“ diameter soo_.,_.,j .ooosand from the equation:CA 19.45 For the gear:-500 = 200 -.ooosMIS = -,0005“dB = 1.8794“ -io0o,_- .,_ ,ootez = 25.72 at 2.060“ diameter “ S;I_T _ _ a_:F. PROF.WAR.)el = 15.53 at 1.9472 diameterand from the equation: Figure 8. TRANSMISSIO
12、N ERROR TEST RESULTSt CA = 20,47 QUIETSET5atpressure will also affect the shrink rateRA_GEI 2S diV STATUS_ PRUSD_,STORED RRS,3O and therefore the resulting part size and,i _ profileshape. 6.0. REDESIGN OF MOLD CAVITY (Plastic: : i Gears).6.1. Measurement of Parts from FirstISd_ Cavity. Molded plasti
13、c parts, made oft_iv i unreinforced material, generally shrinkat a nearly uniform rate. In this case,the material was an acetal (Delrin 500).The original cavity was designed for at specific shrink rate. The resultinglVrml$TRRT= Hz _._ 37.5 Hz STOP=le See M= parts were measured for various diametersX
14、_ 9_8 HZ Y! 1a.6| IVrls THD= 189 such as OD, root dia., rim dia., and hubFigure 9. NOISE SPECTRUM OF IMPROVED dia. The teeth were also measured forGEAR SET involute by adjusting the base diameteruntil the MIS was near zero. This showed4.0. NEW TOLERANCE SPECIFICATIONS. that the base diameter shrunk
15、at nearlythe same rate as all the other diameters.The makers of the gears felt that they The mold cavity was also measured forcouldnt make gears to a q_ality level these various diameters. From thesebetter than AGMA Q6. However, meeting a measurements, it was determined that thespecification on MIS
16、is more of a _riginal parts shrunk more than expected.development problem than a quality This resulted in the gears having aproblem. This involves making the tool higher pressure angle than desired.right in the first place. In this casethe tool is the cavity. 6.2. Cavity Redesign. A new cavity wasde
17、signed and made to the new actualThe specification was changed to AGMA Q7, shrink rate for the material and processto keep the total composite variation being used. This resulted in parts thatdown and avoid a tight mesh condition, were closer to the desired specification.To control noise, a requirem
18、ent of a MIS(fHam) within +/-.0002“ for each member 7.0. EVALUATION OF FINAL PRODUCTIONof the pair was also specified. The +/- PARTS. I.0002“ limit on MIS was based on priorexperience with many other applications. 7.1. Final Involute Tests. Involuteresults of the new parts are shown in5.0. _LANUFACT
19、UEING PROCESS IN RELATION Figure 10 Fxsure II. INVOLUTE TESTS FINAL PRODUCTION48 TOOTH GEAR _ _.(X)04 .RAMG_ 2g d|V STATUSI PRUSED -0.000ERNS=3e 0.072 0.074 0.d76 0.d78 0._ 0._2ze _ _ IlIA = 2.000 _C_ES PUJS_: _ _i FIGURE14.le I :dR,pzv : rate. Therefore a decision was made touse the outside diamete
20、r measurement as acontrol of Mean Involute Slope (SeeFigure 14). As long as the same moldI cavity dimensions are used, this_VrllSTRRT,e,= R.,sr.sHz STOP,leseeH= relationship will hold true. A fixtureleH_.= v,eT.ze.w.= THD,S2S_ was made, that could be used for quickFigure 12. NOISE SPECTRUM OF FINAL
21、measurements of the OD of the gear teeth.PRODUCTION GEAR SET This has now been used successfully inproduction for over two years.Involute Slope. It also shows a goodrelation between outside diameter and_ MIS. Therefore, molding pressure becamea good process variable for control ofthe desired paramet
22、er.“ 8.2 Control of Shrink (Involute) byControl of Outside Diameter. Thediscussion in section 6.0, above, showedthat all diameters, including the base_ diameter, shrink at a nearly uniformWREFERENCES1. Smith. E.E. “The Eelationship ofMeasured Gear Noise To Measured GearTransmission Errors“. AGMA Fal
23、lTechnical Meeting, 87FTM6,Cincinnati, OH, October 6-7, 987.2. Smith, R.E. “Identification of GearNoise With Single Flank CompositeInspection“, AGMA Fall TechnicalMeeting, 8SFTMI3, San Frantiseo, CA,1985.3. Smith, J.D. “Gears and TheirVibration“, Marcel Dekker 983.4. Maag Gear Book, Maa8 Gear CompanyLtd., Zurich, Switzerland, 1990,pages 318 to 320.ACKNOWLEDGMENTGratitude is expressed to the KirbyCompany, of Cleveland, Ohio for the useof data collected on one of theirprojects.8
