For more information"eentaen

Gets the job donein half the time.

A speed gear tlrat's conl'fntionall}' plunge shaved ill 24seconds can now be POWER SHAVED illjust i2 second ...Vast improvements in green having cycle times likethis are mow possible with the breakthrough POWERSHAV[NG'~ proces . The machine platfonn is therevolutionary new Gleason HURTH ZS 130 T. de ignedto greaUy reducethe unproductive "auxiliary" lime.common to conventional shaving machines - whileraising machining speed and feeds La new level. .For example:I. It auto-meshes the workpiece 10 ,

Gold Star Coadnp, Inc.- 2234 S. Dam RoadWest Branch, HI -48661- Phones: (5 17) 3457160(800) 426-2538 Fax: (51T) 3453020

Gold Star Coatings,lnc. 1580 Progress DriveRjchmond,lN 47374 Phones: ('165) 9357424(800) 686]158 FAX.: (765) 9357631NationaJ FAX: (800) 241 ~6616

CIRCLE 128

-

NOVEMBER/DECEMBER 1999

34

BUYERS GUIDE

The Journal of Gear Manufacturing

Definition and Ilnspection lof Prof He' andLead 01 a Worm WbeelNew worm wheel equations andinspection methodsfrom the '0S Gear Lab nH!ardGear Finishing withl alGeometricaUyDefined Cutting lEdge .Skiving can replace grinding for post-heattreatment finishing 24

ICoordina.le Measuring Machines and the Gear IndustryAn overview ofCMM in gear inspection 34

Cov. 1111: 'cow-teS! of IBrown 80Sbarpll. NonbKingl'lOWD. IRI.

IProducts '& ,services Index 40ICompany Index 59

Publisher's IPageThe ueee of OUE customers 7

Revolmi,onsThe Turbinator, Corio lis Drives and 3D Holographic Inspection 111

Iindustry NewsWhat."snew in the gear induslIy? 30

'Technical CalendarMake plans now tor these upcoming events , , , ,32

Ad:verti:se:r IndexTry Rapid Reader Response for nearly instant informatica 33,

19991 AJti,cle IindexA listing by subject of this year's article 191

Product News.New tools for gear manufacturers 83:

Classifli.edsService, Help Wanted and.more B6'

AddendumOne fast gear Boxx............................................................................................. ,

CIRCLE 11)3, GEAR TI:CHNOLOGY

IEDIITORI'ALPublisher & EdUor-in-Cbief

Michael Goldstein

Managing Editor William R. Slot!

Senior Editor Charles M. Cooper

Tedm.ical EditorsRobert Errichello

Don Mc:Vi'lticRobert E. SmithDan Thurman

ARTAn Director Jean BaI'IZ

ADVERlTlSINGiAdvertising Manager Patricia Flam

Advertising Coordinator Donna Lawson

CIRCULATIONCirculation Coordinator Jennifer Beale

INTERNETh1teme~. Edito.r Daniel Gonsiorowski

Gear Industry Home' ~ageT'"SalesPatricia Ram

powertTarmllissiem.com Sal'esAnthony Romano

~ANOAILl PUBUSHING STAffPresident Michael 'Goldstein

V"Ule !President Richard GoldsteinCllnlrulierPatri.ck Nash

Al"Counling Laura ManionArt Consultant Marsha Goldstein

Phllne: 847.,1J,7tm04EmaiJ~people@gearlechllology.com

Web: www.gearll'dlllo!0$Y'Fofflwww.powertransmissum.com

IBPA~V

VOL 16, NO. 16G~~AR TF>CJl"OLOG~r. The JIl'IJrn .. 1 of Go..,.l\Ia:nufacturlJ.~ (ISSN 074HIlI~8) is ~ubll>h

CIRCLE 186

For h'igh volume gear inspectionsystems, you can't beat Moorre.Here's why:

MOORE ExperienceMoore has been designing andproducing high quality, reliable gagingsystems for more than 50 years,

MOORE Technical Back-upMoore's U.S. operations provide over1,000 employees arld a 375,000 sc.ft,manufacturing facility, full engineeringservices, plus additional operationsworldwide,

MOORE High-end ElectronicsMoore Measurement Solutionsdraws on the technical skills andexperience of the Moore Instrumentsand Controls Division, whichproduces state-of-the-artcomputer systems and:software for process control.

--

Aulomal1c fight mesh mspectlon syslem

-

I ._~ ...

P,rchdlo~ ,n!pOdlotl ~Un.ver.sal center cJ,slonce gear checker

-------

MOO.RE's SizeMoore is large enough tohandle major programsand specialized enoughto assure you personalattention.

Call Or fax rhe Moore GeorGaging Experfs Now

Tel: 1-215-646-7400 Ext. 2352Fax: 1-215-653-0347Attention~ Gear Team Mociulor goor blank hrure Manual roll ,Ionel

[MPOREl"rhe' Measurabl'e Dilferenfe"

Moore Measurement Solutions1.201 Sumneytown Pike, Spring House, PA 19477www.moore-solutions.com/mms

CIRCUE 106

http://www.moore-solutions.com/mms

-

The Success of 0 r ustomematter what bu ine s you're in.

O you need customers. Moreimportantly. you need cus-tomers who can and want topay for your good or services. U'sin ourbest interest to do everything we can tomake sure OUj eu tomers are uceessfulwith the product or service , they buyfrom I.l I as I believe that our wages arepaid Din by OUj companie , but by theircustomer.

This is true of any bu iness-for pub-lishers as well as gear manufacturers. Forexample, if an advertiser places an ad andforgels important information. like hisphone number, it's more than beingnlc~it' good business-when I informhlm that it would be w:i e to include it.

U the advertiser place an ad and hegel no re ponse, whose ffllllt is it? Doeit matter? From my perspective, whatmatters j that he's not going to be uc-ces :fu] wi!h our magazine, and he's notlikely to advertise again. What I want ismore cu tomers, not fewer, so if there',anything I can do [0 make him more sue-ce sflli. it's in my be t interests to makesure he doesn't fail

Sometimes if easier to see thesethings when you're the cu tomer. I'verecently had an experience I'd like tohare with you.

Not long ago we purcha eda newcomputer EoI' our office. directly fromMicron Technology, Inc., one ofAmerica's large t mail order computerupplle:rs. We have several other Micron

machine . and each of them has a1wa)'sperformed well.

When this particular machine arrived,we found that for some reason it wouldnot connect to our Novell network Wecouldn't understand the reason. becau eevery other machine we had ever bought.including our other Microns, attached tothe network with no problem, right out of

the box. Of 'course. we consulted Micron.and they were extremely diligent in tryingto help us olve OUj problem. They had usend the machine back to them, where

they worked on it, swapped out orneparts, and tested it.

However. when we received themachine back ..it still didn't work. Microncontinued to work. with us to tty to olvethe problem, and eventually we were putin touch with one of their Novell-certifiednetworkengineers, He, 00, was baffled,From what he told us, the machine wassuppo ed to do what we wanted and need-ed,and there snolild be no re on why itwould not work. for us.

Thi went on for some time. inally. [received a letter from Micron's "Office of'the President," which informed m thatMicron was washing its hand of theproblem. that Lhey were unable to support3m party software (which in our casemeant Novell). Despite all the efforts ofthe customer ervice people and the engi-neer who tried to help, us along the way.this one letter made me feel as though thecompany di.dll't care that. their productwasn't doing what the)' or I expected il. todo. ~ responded with a letter of my own,expres ing the e feeling and asking whatI wa uppo ed to do with the machinenow. Wa Ijust out of [lick?

Micron has yet to respond. Right now.the machine is itt.ing in a corner o:f ouroffice, back in its box, and it ha n't beentouched for month .

The reslly untortunate thjng is !halexcept for!h offic-e of the president.Micron seems to have its act together. Infact, some 'of their sales literature soundsalmost like I could have written it myself.The "About Micmn" section of the compa-ny's Web page begins: "MicronElectronic is preparing for . uccess, Yoursuccess," The ection titled "Vision," endswith the statement" ... ultimately. your sue-

cess is our success ... and we wantro win."How is it. then, that ] fell through the

cracks?'

11seems to me that it's an awful wastefor Micron to have spent. an that time andmoney trying to help me only to havetheir "Office of the Pre ident" decide thaidoing bu ine with m wasn't wonhtheir time anymore. With one decision,the company' management not onlyundid all the goodwill. that it employeeshad worked so hard 10 build. but they'vecaused irreparable damage to a relation-ship with a customer who, under othercircum lances, would have bought fromthem again,

I understand the concept of cuttingyour los es, of not throwing good moneyafter bad. Perhaps there are mitigatingcircumstance about which [ am unaware,AU m know i that [ have a. compllter thatwon't work for me, and J feel as theughI've been Ii'll down. I wonder if Micron'employees feel the same way.

Michael Goldstein,Publisher and Editor-in-Chief

NOVEIM!BE!lIDIEC'EJoIBER .UI '1

KAPP and NILESare manufacturers of gear and profile grinding machines for the automotive, aerospace

and commercial industries: innovative - reliable - efficient. The compound pinion at right used for powering

heavy equipment. is ground on a KAPPVAS machine to increase load capacity. Call us for details.

Representing KAPP. NILES and KAPP TECH:

KAPP SALES & SERVICE LP, 2870 Wilderness Place,

Boulder. CO 80301, Phone (303) 938-9737 Fax (303) 447-1131CIRCLE 194

KRPPSALES & SERVICE

HOfter Mascmnenbau GmbHIndusbiestr. 19D76275 Ettlingen/GermanyTel. +49 1243 5990Fax +49 1243 599165

MEGA 1250 16002500

Gear dIarnd"er up to3500 nun (l38j0rtndblJ Dlke 1000 mm (401HelIx ar1Jle 35Module 2 - 30 (Dr 13 - 0.85)

For moreinformationcall or write:

CIRCLE 112

_-------------~!R'~VOLUTIONS La~~\:I~~~:i~:t~~iHI'oln:

The machine is called a streamliner. Itis thirty one feet in length and three feelwide. It weighs 3,400 lbs, and looks likesomething out ofa science fiction movie.It's the Thrbinator HI .andit's themachine that Don Vesco used when hebroke the land speed record for wheel-driven, gas turbineautomobiJes (asopposed tolbru t vehicles that use jetengines to reach speeds over 700 mph)out at Utah's Bonneville Salt Aats thisyear. Vesco's record-breaking speed was4]7.529 miles per hour. To accomplishthis, Team Vesco, put an Avco LycomingT55L IlA SA gas turbine helicopterengine in the car ..The engine generates3,750 horsepower at [6,000 rev/mille-more than enough to push the Turbinatorill (0 record- breaking speeds. In theory,the Lycoming could push the car past 600miles per hOUT, 1,00 miles per hOU1 ormore over the tires' top, rated speed!

Of course, peed isn't everything ..There is also traction.

At speed, the tires are almost neverfirmly all the ground at the same time.According to Don Vesco, the driver,mechanic and co-designer of theTurbinator, the tires jump and skip oversections of the l l-mile course. This sub-jects the driver to vision-blurring vibra-tions and the drive train to varying loads

as the car builds up speed and races forthe finish line. That is in addition tothe9,000 lbs, of gear-stripping torque com-ing out of the Lycoming Turbineengine,That's enough to give any transmissiondesigner night sweats.

The solution that Vesco and his geardesigner, Bob Hodgkinson, came upwith was to eliminate the transmissionentirely. Instead, they built a gear reduc-er with three gears made fromCarpenter's Aennet 100 steel. This, cou-pled with the locked 1:1 from and reardifferentials, means that the Turbinator'sdrive train is actually a direct drive sys-tem. "We wanted to make the system asbullet proof as possible," Vescoexplained. "The gears are AGMA 12andwere cut and ground from a single bar ofAennet 100 with the shafts integral 'tothe gears." This allowed them to' increasethe strength of the parts and avoid theproblems associated with shafts flexingand moving in their bearings.

What made this possible was theengine itself. According to Vesco, it is theturbine engine's blades, not the engine'spower plant, that tum the driveshaft andthe wheels. "It's basically an engine witha hollow output shaft running throughthemiddJe," said Vesco. "The output shaftcomes out the fron: and ha four fanblades in the back. The exhaust from theengine turns these ran blades, The fan

Don Vesco, and Ule Tarbinlllor m..Courte.sy of Team Vesco.

Welcome to IRevol'utions, the co/~umn that brings you 'he 'atest,most up-to-date Bnd easy-to-reldinformation about the peopleBnd technology 0" the gearin.dustry. Revoludons wel,comesyour submissions. PleBse sendthem to Gear Technology, P.O ..Box 1426, Elk Gro", Village, Il60009, fax (Sf7) 437-66tB or ,-mailIpeople@geartechnology.com. Ifyou'd like more information aboutBny of the articl,s th,t IppBBr.please ci~cletbe appropriate num-.beron the Reader Service Card.

blades, astbey tum, drive the car." Thimeans that the engine it elf acts like atorque converter in an automatictrans-mission in that it bui Id UP. yet neverprovides full power to the wheels due toa slight energy loss in the turbine itself.Therefore, it takes a little time for thewheels tocatch up to the power plant.

The wheels aren't the ollly thing thathave to catch upto the power plant. It'taken Vesco himself a few run to getused to driving !:hat fast. "At. first ~ wasamazed, five miles gone and I didn'trealize it," he said. "Now, I'm waitingfor it, My mind is going as fast as thecar;" Today, that speed is 417.529 mileper hour, Next year, 500 miles per hourwill be within their grasp. "With. ourhighest ratio gear reducer and enoughtrack, the car could probably go as fast as700 miles per hour," said Vesco. "Butrignt now, that's not realistic." Perhaps,but it is only a matter of time.

Circle 2St

H!olographicMeasuremenlls i:nl301

Non-contact measuring sy terns havebeen electro-optical or laser-ba ed innature, the first taking a digital image ofthe part under inspection and comparing

NOVEMBER/DECEMBER 1nl 11

mailto:Ipeople@geartechnology.com.

it to a digital nominal part, the secondusing lasers as a kind of touch probe.Both are precise methods, but the folks atOptimet believe that they have come upwith something better ..They have intro-duced holographic measurement to theart of gear metrology.

Optimet, a division of OphirOptronics, Inc., has just developed a newway to precisely measure parts. Calledthe Conoprobe 1000, the unit is a gener-al purpose, non-contact measurementprobe that uses the concept of conoscop-ic holographic measurement to createthree-dimensional digital images of theparts it measures quickly and fromremarkable distances.

What is Cenoscopic Ho.logr,aphy? Inregular holography, a three-dimensionalimage is formed when an interferencepattern is created between two coherentlight sources, such as lasers. The beamsfrom these light sources, called theobject beam and the reference beam,travel at the same speed but they followdiffering courses. This creates what iscalled the Gabor Zone Lens (namedafter Denis Gabor, the Hungarian physi-cist who discovered holography in1947) and the image.

Conoscopic holography is slightly dif-ferent. It uses the ordinary and extraordi-nary components of a single coherent lightbeam passing through a urn-axial crystalto create the hologram, This conoscopichologram has fringe periods that can beprecisely measured.

The Co!!oprobe 1000. Courtesy of OptimeL

12 GEAR TECHNOLOGY

REVOLUTIONSThe Technology. This method uses

concentric optics that function regardlessof their position to key optical elements,making the system flexible and ruggedwhile maintaining repeatable precision to1I8000th of the working range. The scaleof the measurement, from sub-microns tometers, is adjusted by changing the objec-tive lens on the probe. And because theprobe is collinear, changing over to bend-ing optics will permit measurements to betaken around blind comers.

The new system is also surface inde-pendent, meaning that it can create holo-grams from a greater variety of surfacesthan previous non-contact methods. Thisincludes very shiny objects as well asthose with wide variations in reflectivity.It is also capable of working very closeto grazing incidence, a mere five-degreesfrom normal incidence in all.directions.

The Conoprobe 1000. Designed tobe integrated into existing measurementsystems, The Conoprobe 1000 is capa-ble of taking up to 700 data points perminute while the probe is in motion.This permits the unit to develop preciseholographic images of virtually any partincluding machine parts and tools, plas-tic and rubber industrial molds andcomponents, auto parts, electronic partsand more.

Circle 252

The CarioUs DriveIt all started with an observation made

over a century ago. If you lay two coinsof the same size next to each other on atable and then roll one around the other,the coin in motion will rotate 720.Why? The answer to this question, calledthe "Two Penny Paradox," is as valid forgears as it is for pennies and, accordingto Ken L. Baker, a design engineer forFleetwood Systems, Inc., of Romeoville,illinois, it is the basis for the Coriolisforce, the principal upon which hisCoriolis drive works.

According to Baker, the way that theCoriolis drive relates to the two pennyparadox is like this: "A wheel turnsaround another wheel and it goes around

has travelled and once because of theshape of the path." This is an example ofa law in physics called conservation ofangular momentum. The penny has tomove an equal distance in two direc-tions; therefore it has to rotate twice.

The same principal makes a rotatingdisk wobble as gravity pulls it down."Roll a coin across a table and observeits motion after it begins to topple," saidBaker. "It's propelled by linear momen-tum and rotational. inertia while underthe influence of gravity. it swerves intoan ever-tightening spiral course andeventually starts wobbling around a sin-gle point. If you look at it closely, you'Ilsee that it is rotating backwards."

Geometry can explain part of thephe-nomenon. Due to the angle of inclinethat the coin makes with the table, theradius of the path (and therefore its cir-cumference) is shorter than that of thecoin. "For each trip around the path,"Baker explains, "the coin is required tomake less than one complete rotation. At1: I, no rotation at all would be required,but at greater differentials the rotationreverses for exactly the same reasonsthat cause the Two Penny Paradox."Prom that coin wobbling on the table it isjust a short leap of the imagination toBaker's machine.

Baker's Coriolis Drive takes advan-tage of the difference in rotation betweenthe wobbling action and the revolvingaction to act as a gearless speed reducer.With Baker's device, when you start itspinning, all it does is spin. But whenyou start. it wobbling, some of that spin-ning motion is converted into that wob-bLingaction. "For example. if you have itrevolving at a speed of 100 rpm and itbegins to wobble, and that wobble caus-es a 1 rpm precession motion, the origi-naJ rotational speed will drop to 99rpm." Taken together, the two rotationalspeeds still equal 100 rpm. This satisfiesthe law of conservation of angularmomentum. What determines theamount of precession is the amount ofwobbling. "The greater the force thatwants it to fall (wobble)," explainedBaker, "the greater the force that wants it

two times, once because of the distance it to precess."

REVOILIUTIOIN:S II.__ - ~

Where' You Want To BeAGMA has just released a new 14-

minute video designed to introduce stu-dents to the gear industry. The video,called "Where You Want To Be: AnIntroduction to the Gear Industry," was acooperative venture between AGMA'sEducation Council and the AGMAFoundation. The project was undertakento increase awareness of the gear manu-facturing industry as a career option for

Rather than create the wobble withgravity, Baker chose to do it mechanical-ly, to force the wheel to tilt. "The tiltwill, itself, pull enough energy out of thewheel's rotation to cause the wheel toprecess," said Baker. "I'Il use that ener-gy. That's my gear reduction. I'm lever-aging one spinning motion againstanother spinning motion the way thatgears leverage the short radius of onegear against the long radius of anothergear. Gears use geometry, and what I'musing is physics. But both of them comefrom the same mathematics. They comefrom the same science."

Baker has developed a workingdemonstration model of his Coriolisd.rive that he is currently in the process ofrefining. Apart from bearings and aframe, the machine has only four movingparts-a drive shaft and disc mountedoff center at an incline, a rotor on a fixedaxle. a flexible universal-type couplingand an output shaft.

According 10 Baker, 'The offset ofthe disc is what maintains a fixed angleof precession, as the central perpendicu-lar axis of the disc is coaxial with that ofthe rotor. A projected extension of thisinclined axis would intersect the majorinput/output axis atlhe exact center ofthe universal joint. Thus, by rotating theinput shaft, the rotor and its axis areforced to precess about the major axis ofthe universal joint, This precessional.motion induces rotation into the rotor,which causes the output shaft to rotate inthe same direction and with the sameforce."

While Baker freely admits thai. thereare a number of applications where aCoriolis drive would not be a viablealternative to gears, he also holds thatthere are quite a few areas where hisinvention would, as he puts it, "outshinea gear motor and perform the same taskmore efficiently."

These applications include variablespeed, low-friction motors and transmis-sions for applications ranging from smallpower tools, Iawnmowers and pumps tomotorcycle and helicopter engines anddiesellocornotives.

CWcle 253

high school and trade school studentsand their families. Twenty-eight gearmanufacturers contributed to the AGMAFoundation in support of this effort.

The video presents gear manufactur-ing as an industry featuring modem.dean, high-tech facilities where employ-ees will be trained to lise the latest incomputerized and electronic machinery.It also describes the gear industry asbeing competitive in pay and benefits.

PROCESSInspectionSince 1936 ITWhas provided the gear indusLrywith gear inspection devices. Put your trust in

the people who invented the process.

PRODUCTS AVAILABLE: Manual double flank testers for coarse pitch. Manual double flank testers for fine pitch. Computerized double flank testers for

coarse pitch. Computerized double flank testers

for fine pitch. Dimension over pins or balls, Automatic in-line gauges,

Model 2275DOf'Dimension over I

Pins or saus

No matter what the application; coarsepitch tine pUch, externals, internals,

shafts, metal or plastic - we lookforward to working with you.

IITW Heartland1205 36th Avenue West

Alexandria, MN 56308 U.S.A.Ph: (320) 762-8782Fax: (320) 762-5260

Email: Itwgears@rea-aJp.com

CIRCLE 1119NOVEMBER/DECEMBER 1 BBB 13

mailto:Itwgears@rea-aJp.com

_------------- REVOLUTIONS _geographically distributed across thecountry in both rural and urban areas, andoffering positions ranging from machin-ing to design 10 management. Finally,gear manufacturers are shewn as compa-nies where a positive attitude will buy theoppertuniry to work in a dynamic andchallenging atmosphere producing thehigh precision products that keepAmerica competitive and on the move.

Copies of the video will be provided.free of charge to AGMA member compa-

nies, and additional copies can be pur-

chased from AGMA. Brochures, whichsupport the message of the video, arealso available, Gear manufacturers areencouraged to contact their local second-ary and trade schools and provide copiesof the video and brochures to the guid-ance staff. To obtain copies of the video,and brochures, contact AGMA at (703)684-02] L

Circle 254

C C Sipline! iRolI,er 1 Precision Splines to AGMA Class Q1:2 5 to 7 eNC .Axes Quick Change Over S - t: Wide FlexibilityI. Up to 3 I:lifferent SplineS' in one Set-Up Rolls Tlirough-H'ard'ened SteB'1

and Hollow Shafts

14 GEAR TECHNOLOGY

IERNST GROB AGCold-forming machines

1'jr:...CH-S708 MannedorflSwitzerlandPhone +41-1-922 77 00

Fax +41-1-922 n 88Intemet: http://www.emstgrob.com

E-maH: info@ernst-grob.com

Cal'edonlan Midwest Sales,. Inc.5497 Daniel Drive- 'Brighton,Mil 48.1149069Phone (aW) 227-3977 Fax (810),227-4771

E-mail: dempster@ismi.netCIRCLE 11:'6

C'orrectionWe apologize. for; and wish to correct,

the following errors that appeared in theRevolutions article featuring FalkCorporation in the September/October

1999 issue of 'Gear Technology.The name "Fimeston" should be

spelled! "Fimistoa," The OEM for the

mill located at Fimiston Mines is fFEMinerals of Australia. The flanges ofFall's large gears are locked togetherwith tapered steel dowels, not lockingpins. The pinions mating with theselarge gears are finished to AGMA 12levels, not the ring gears themselves,Ring gear segments are finish cut toAGMA 10 tolerances with average tooth

finishes falling around 63 to 100 R_MS(not RMF). Average tooth finish isrequired because there are areas that

could get as high as 125 RMS and typi-cal consultant specifications require amaximum of 125 RMS .

Again, Gear Techn.ology apologizesfor any confusion, Inconvenience or con-

sternation these errors might have

caused. 0

DO YOU IHAVE.AN INTE.RESTINGI

GEAR STORYTO TELL?

SUBMIT YOURI'DEA TO

REVOLUTIONS,P.O. BOX 1426.

ELK GROVE VILLAGE,IL 60009 (USA)

OR CALL(847) 437-6604 ..

TaU U. What You Think ...If you found these Revolutions of interestand/or useful, please circle 22D.

mailto:info@ernst-grob.commailto:dempster@ismi.net

Ipowie!r I-teglratedPOp:-UpITMliiftlFlotate IinductionHeat~1ireating System

Radyne's PO\IVer Integrated Pop-upTI.IIheat-treat oenter lis a self-contained.system for hardening and temperingcomponents In a lift/rotate, sub-merged quench method to meet thespecific needs of the heat treater. A.user-friendly maclhlne interface paneland PLC control enable quek andeasy setup and operation. Anintegrated. modern, effiCient.transistorized inverter IpovYer supplycan match a Vllide variety 'of heatingcoils 'VVith easyto-change tuningcapacitors and a multitap outputisotation transformer.

The lift.actuator assembly includesa baD bearing linear way mountedundera stainless steel sink throughdouble lip WENe seals. A chrome-plated stainless steell spirndle ismounted on a tapered roller bearing,enclosed in a steel housing. The lliftmechanism aJlows load/unload,heat and quench positiOns.

The P'erfect IntegrationlThe combination of Radyne's PbvYerIntegrated Pop-upnA and APEX QATI.IIQuality Assurance systemrepresents the latest in inductionheat-treating technology.

RADY E.inJ1DV'tors in Indallli'on Heltin,'-800-:236'..,8360211 W. IBoden Street!Milwllukee. WI 53207. U.S.A.(414) 48,1-8360 FlOC(414) 481'8303email: rady1le@execpc.comhttp://www.radyne.coffil

CIIiICI.E 139

http://www.radyne.coffil

9lit -!7nkmationa/

The Inductoheat Group is hosting its9th lnternat.io,.nal Induct.ion Heating SeminarMay 10-12,2000 at the Hilton ClearwaterBeach Resort in Clearwater Beach" Florida.

The world's largestgroup of experts from North America, Europe,Australia, South America and Asia will present information on NEWprocesses and developments in induction heating.

Breakthsough in the induction hardening of crankshafts Innovations in induction hardening of camshafts, gears &criti-c~doomponeats Obtainin,g desired metallurgical properties Cracks, shape & distortion control Induction tempering &stress relieving Ferrous & non-ferrous metals Billet/bar/wireheating S,lab/strip/plat,e heating Tube & pipe heating Cellular manufacturing designs- New ceil design technologies Diagnostics, monitoring & QA Brazing & soldering technology Failure analysis IGBT & .MOS FEr transistorized power supplies Load matching Advanced induction heating techniques" & more

New Solutions to Current NeedsBring your questions, concerns and induction experiences to worldrecognized experts and end-users and get the advice and answersyou're looking for.

Two-and-a-Half Day Seminar Registration/reception Tuesday evening Forma] presentations Wednesday & Thursday Roundtable discussions Friday until noon

It's Not Too Early to RegisterBe sure to register via phone, fax, e-mail or our website to guaranteeyour spot in the seminar. The seminar fee is $195.00 per person untilApril 10, 2000. Regisrration after the 10th is $250.00 per person.Please make checks payable to Inductoheat, Inc.

eN'DUCTOH:EAT'

GROUPCIRCLE 161

.'INDUClOHEAl LCO RADYNE DlWlIlU:rulUl' OC ALPMIl. IHS IIIIIIG INDUCTOHEAT

The 8th International Induction Heating Seminar,held in November, 1998,featured guest speakers andattendees from 19 countries,

Heating Seminar will be held at the Hilron ClearwaterBeach Resort in Clearwater. Florida,.

Contact:Laurie TaylorIN DUCTOH EAT, INC.32251 North Avis DriveMadison Heights, Ml 480711-800-624-6297Tel: 248/585-9393 Fax: 248/585-0429e-mail: lItaylor@inductoheat.comwww.inductoheat.com

mailto:lItaylor@inductoheat.comhttp://www.inductoheat.com

_------------INSPECTIONI------------.Definition and Inspection

of Profile and Lead 01"a Worm Whl,ee,l

lntroductionTraditionalily, profile and lead

lnspeetions have been irtdispensablepor-

Lions ofa standard inspection of an invo-

lute gear. This also holds true for theworm of a worm gear drive (Ref. I), Butthe inspection of the profile and 'the leadis rarely performed on a worm wheel.

One of the main reasons is our .nabilityto make good definitions of these twoelements (profileend lead) for the wormwheel. Severn re earchers have pro-posed methods for profile and. leadinspections. of a worm wheel using CNCmachines or regular involute measuringmachines. Hu and Pennell mea ured aworm wheel's profile in an "involute"

section and the lead on the "pitch" cylin-d r (Ref. 2). This method i applicable ,toa convolute .helicoid worm drive with a'crossing angle of 9Q,o becausethe wheelprofile in one of th offset axial planes is

rectilinear. Thi traight profile gener-ates an involute on the generated wormwheel. Unfortunately, becau e of the hoboversize, me crossing angle between thehob and the worm wheel a:lways deviatefrom 90" by the wivel angle. Thu ,'this

Dr..D,onaldilR. H!ouser ,and Dr: XiaagenSumethod can be implemented only

approximately by ignoring the wivelangle. Another shortcoming of this

method i thai there is only one profileand one lead on each flank, If thecanned points deviare from this curve, it

produces unreal profile deviation. Ocnuedi cussed profile inspection using a pro-file checking machine (Ref. 3),

If the swivel angle is not ignored, tileinvolute profile used in Reference 2 doe .not exist. and the profile equation devel-oped in Reference .3 does not apply.Thus, profile and lead inspection is 1'101often performed to qualify a wormwheel. To verify the compatibility of thewheel. 'toolh with that of a mating worm,the wheel is assembled with a masterworm ona te L rig and a roUing test isperformed ,(Ref. 1). The contact patterndeveloped by painting the worm is usedto judge the quality. This kind of rollinglest works well in terms of functionality,but it doe not give enough quantitativedata to the designer and the .hobbingmachine operator. In other word , therolling te l re ults cannot be fully usedfor qua:li.ly improvement.

------ - - -- - ---

NOMENCLATUREE Center distance between the hob/worm axis and the wheel axis while meshingXI ;; IxlY zl.worm surface point vector~ = [~. Y2' ~IT, wheel tooth surface point vectorm21 The gear ratio (number of worm threads I number of wheel teeth!n1 = InK!' nJil' n1l, normal vector at a worm surface pointt ;; -P~I'equivalent translation displacement ,of the wormu Surface parameter on the'worm thread in radial directionV,1Z The relative velocity between part 1 (hob/worm) and part 2 (wheel) while meshinga Profile angle ofthe grinderr Crossing angle between the worm axis and the wheel axis while meshing~l The worm rotation engle while meshing8 Surface parameter on the worm thread in circumferential directionP Screw parameter of the hob/worm, equal to the lead divided by 21t

Z{,

(a f Rotating wonn

-Zf',

IFig. l-TwlI' equiv.erd meshinll matio I 01 IIwonn IIlIlr dr,iv .

IDr.ID,onald R. Houseris director of both th Ce/l/I!T of AU/OlIIDlilleResearch and t~ Gear Dynamic and Gear NoisResl!Qrch ~borQ/(Jry aJ Ohio. 11l11! University. Hisresearch is directed toward Ihe reduction of gearno; e Ihrough modificarion 10.Ih gea, loath sur-fact duigTl. Research is al {}oriented IOwaN /hemeasuremellt of dynamic and SIalic transmissionerror of gears as well as automotive noise mea-surements for SOl/lid quality (!VQ/uo.rion.

D,r. Xjaogen Sureceived his doctonue !ITJm ,ohio tat Unilfl!r.rilyin June 1999. His principal Ilrta ,of" eQl'Ch igeQr ill pection and rever: e tllginuring. Heworked as an engineer and teacher for Sl'Vlmyears before coming 10 me U.S. and now is;>mp/o.vedin hina by Ihe Cope/mId' 'orporation:

NOVEMBERIC C MDER 18119 1'1:

_-------------INSPECTION1i------------_In this article. an explicit equation of

the generated wheel tooth is derived. This Iequation applies to any type OJ singleerweloping WOn-II gearing. The ero singangle between the hob/worm axis andthewheel axis is 1101 limited to 90. Thus, thisequation can also be used for non-eight-angle worm gearing. Based on the explicittooth equation. profile and leads of a wormwheel are redefined, The new defin.itionsare comparable to their counterparts of 311involute gear. There are ,311 infinite numberof profiles (at different "face po iuons")and leads (at different "diameters"), asthere are on an involute gear. All of the pro-file and lead curves can be expressedexplicitly and can be readily measured by aCNC measuring machine, Inspectionexamples are given (a CMM is used for themeasurement) to illustrate the application.

! Explicit !Expression of the GeneratedIF----""""---;:::==~::; I i Worm \Vheell Tooth

The geometry of a generated wormwheel is very complex. An explicit equa-

I Lion of the wheel tooth surface has 110tbeen found before. The traditional. 'expres-sion of the generated worm wheel tooth.as presented in Reference 4, is in the formof:

wolll1{gear edges- u lines (contsnl 91- Dlines [ccntant J..lI

e

tip

(bl Generated gear surface 1 2

Fig!. 3 Multjp'I', IProfile3i and I'eads.

(II worm grimfing setup

z,

(bl worm grinder

Fig..4-Gr,inder for ,ZK-type!oJWOIIII.1!8 GE.AA TECHNOLOGY

From Figure]. the sam point and itswhere Xl = [xl' YI' zif is the urface point normal can 'be expressed in coordinatevector and nl :::: [n.d, nil' n)T is the corre- system SfI aspending normal vector of the WOITn thread

X = X (1,9, 411)Subject to f(u,8~ 4'1)= 0

The oonstraintf:::: (u,9. 4'1)i the equa-tion of meshing for cutting. The commonlyused equation of meshing for worm gear dri-ves was proposed by Litvin (Ref. 5).11 is:

surface in a. coordinate system attached tothe wonn itself (See Fig. 1). The expressionXl ::::Xl (1,9) of the worm thread of vari-ous types of single enveloping worm geardrives can be found in Reference 5.

Equation 3 has three parameters: u; ()and ;1" Here u. and 9 are the two urfaceparameters of the meshing worm surface

and variable 4'1 i the rneshjng parameter(the worm rotation angle in this equa-tion). It is difficult toexplicitly expressany variable among u, (J and 4'1 in termsof the other two. With this equation ofmeshing, each point o.n the wheel tooth isassociated with three variable that areconstrained by the meshing equation (Eg.3). This brings much inconvenience whenperforming surface description and ur-face computation (Ref. 6).

Our new expres ion for the generatedworm wheel tooth is made possible wilhII: new equation ofmeshing for worm geardrives. The new equarion of me hing isbased on an ob ervation that the rotationof tile worm abou: it); axis IIlQ,}' be kine-matically reptaced by a transkuion alongirs axis (See Fig. 1b). The relationbetween the rotation angle 411 and 'thetranslation ,t is

(4)

We will derive the equation of mesh-ing withthe tran lating meshing motion.The general equation of meshing for alltypes of gearing is

D. "v.,!2",QI r

(5)(l)

(2) where i represents the coordinate sy tern.For convenience, the fixed coordinatesyslem SfI ( liIown in Fig. Ib) is u ed.Coordinate sy tern SfI coincides with thecoordinate ystern 51 attached to theworm when the worm ha zero transla-tion. In coordinate system SI'

Xl= [xl' Yl' zd'11) = [1l.d.I1."J' II.Jf

(6)(7)

XfI= [x)'l'}'n'~].T:::: [xl' Yl' 'I-P4'IY (8)nfI:::: [fl.ifJ' Ilifl' "dllT:::: (/lXI.llyl' f1z1f (9)

The relative velocity at the meshingpoint in coordinate system S'fI is v}l '" v~

2 - d- vfI' anVf11::::(0, O. -P4'I'jT

k

INSPECTION: _

m21 tfi'lsiny T112I lfJ'ICOSYIYI Zl - PI

~!I

Ii

II!,.From equations 5, 9 and :10, we get thal::~:::::::~::::~ ~,+ E) ICOS'j'l'ly\ + [p/m21 - (Xl + E) sinn n~1 ~ l~) ',I

Thus, the meshing parameter lfJ'j can beexpre sed explicitly by the two surface i

!parameters, u and B, of the generating ,

I,!(xI + ) cot/fl~'l+ (plm21/smy- (XII + n;:1 + (

_IINsP,E'cT'ION IIII

,and ills mad'e in AMERICA'!

that allow them to be defined at multiple

locations aceess the tooth surface. For aninvolute spur/helicaJ gear. the tooth sur-face can be written as:

X"" X(E:./J (14)

where E is the roll angle and/i.s the faceposition parameter, A pair (E, f) deter-mines a point and its normal 01'1 a certaintooth flank, The defined profiles and

AfW Systems Co. announces that it is, now amanufacturing source of spiral gear roughingand finishing cutters and bodies.

We also can manufacture new spir,sllcutter bodies in diameters of S' through 12'at present.

AfW can also supply roughing andlinishingcutters, hardware and replacement parts formost 5'-1.2" diameter bodies.

Whether it's service or manufacturing, con-sider us as an alternative source for replace-ment parts and hardware as well as bodiesand cutters.You'll be in tor a pleasant surprise.

NEW! Straight Bevel CuHe,",s.

lead of the involutegear correspond to

the slines and the/lines" respectively. Ifthe parameter f is fixed at/o' the curve X= X( E, fo) represents a profile of the gearat face position /0; if the parameter ,I: isfixed at EO' tile curve X "" X( Ea, j) repre-ent a lead of the gear at roll angle Eo.

An equation (B) of the generatedwonn wheel tooth similar to Equation ]4

for the tooth of an involutegear has beendeveloped:

Royal Oak, 'Michigan 48067Tel: (248) 5443852. Fax: (248) :544-3922

201 GEAR. TECHNOLOGYCIRCLE 1111

~ = Xl. (14, fJ) (15)Here. we define the profiles and the

leads of a worm wheel in a format similarto that used for an involute gear. FromEquation ]5, if the parameter 9 is fixed at90, the curve ~ = ~ (14, 90) traced out. bychanging u is defined as II profile of theworm wheel tooth; ifthe parameter u isfixed at /,10' the curve ~ = Xz (140, fJ)traced out by changing e is defined as alead ..The profiles and the leads defined ill

this way nave explicit. mathematicalexpressions, so the profiles and the leadscan be proyammedand mea ured withCNC mea uring machines.

The profile defined by X2.'(u, (0) isnot. aplanar curve. It. runs across theloath surface from the root area to the tiparea, and this profile is called a profile atparameter 90, The lead defined by X2(uO'fJ) does not lie on one cylindrical sur-face. It runs across the tooth surfacefrom one edge to the other. and this leadis called a lead at parameter "0' Thereare an infini.te number of profiles and!leads on each flank (Fig, 3). Any surfacepoint is the intersection between at pro-file and a lead, and there is one profileand one Lead passing through the pitchpoint of the worm wheel.

Take ZK-type of worm gearing as an

example. A ZK-type of wonn is groundby a biccnical grinder (Ref. 5), as shownin Figure 4. Parameter 11 is the distancefrom the apex. ofthe cone to a point onme grimier profile ..The u lines (profiles)and the e lines (leads) on the hob/wormthread and the generated worm wheel areplotted as in Figure 2. In this case, thebob/worm profile is not the axial sectionof the .11Ob/wonn thread. 1l is the tangency

line between thehob/wonn grinder andthe hob/worm thread. The hob/worm lead

is the helix line. Because of the one-to-one mapping of points on the hob/wormthread [0 tile points on the worm. wheel

tooth defined by Equation 13,. ahob/worm profile generates a wheel pro-file" and a hob/worm .. leadgenerates awonn wheel lead. Obviously, the profiledeviation and the .Iead deviation mea-

sured all the whee] directly reflectthe

errors of the hob and the hobbiiog set-tings. Thus, the inspectien results can be

_------------INSPECTIION------------_more efficiently used for later qualityimprovement.

Figure 5 shows the profiles (.1 lines)and the leads (,0 line-s) of a ZA-type ofworm gearing. In this case, thehob/worm profile is the axial section ofthe hob/worm thread, and the physicalmeaning of the parameter Ii is shown inFigure 6.

Inspection of P~o6i1e.Lead and Topograpby

The profiles and leads defined abovehave explicit equations, and a CNC mea-suring machine can be programmed tofollow a profile/lead trace. For the pur-pose of inspection, the measured tracesare compared with their theoretical posi-tion to produce the deviation charts. Thedeviation chart of the profile X2(u, (0) isplotted against the parameter u, and thedeviation chart of the lead ~(uo' e) isplotted against the parameter 9.

There are two ways to perform theinspection of a wheel First, the wormwheel can be inspected against the mesh-ing process (the worm design and themeshing setup). The actual wheel toothsurface is compared with a virtual wheel,which is conjugate to the worm part.Large surface deviations are expectedbecause of the difference between thehob and the worm. The deviation of the

positions" and leads at different "diame-ters .." Two different presentations can beused: the chart is plotted over a grid ofparameter-s x parameter-Per plotted overa grid of radius x face position.Example of Worm Wheel Inspection

A 30-tooth, ZK-type worm wheelwas inspected. The measured wheel toothwas compared with a virtual wheel toothconjugate to the mating worm. Figure 7shows profile traces for both sides offourseparate teeth. The inspected teeth are thel ", 81b, 16th and 23rdones. The root reliefshows up dearly, but the left flank hasless tip relief than the right flank. A pos-sible reason for this is nonsymmetry ofthe hob grinder, but it may corne fromtotally different aspects, such as an orien-tation error of the wheel.

Figure 8 shows lead traces for bothsides of the same four teeth. The endrelief introduced by hob oversize is con-siderable. With the decrease of hob over-size, the amount of the end relief isexpected to decrease. The swivel angleadopted for wheel hobbing has a directimpact on the slope of the lead ..The accu-racy of the middle face position mayaffect the slope of the lead trace. Thisaccuracy is discussed in the next section.

Figure 9 and Figure 10 show the mul-tiple profile and lead traces of one tooth

profile tellsthe amount of tip relief and of the same wheel, The profile tracesroot relief, while the deviation of the leadreveals end relief introduced by 'the hoboversize. Second, the wheel can beinspected against the hobbing process(the hob design and the bobbing setup).The deviation caused by the differencebetween the 1I0b and the wonn part willnot show up as surface deviation.Because the hob shape changes due to.resharpening and/or wear, the designdimensions of the hob are used. Themeasured deviation would nOW revealthe difference introduced by the decreaseof the hob oversize. Very likely, the leadlias a negative crowning rather than pos-itive. This inspection may give a goodidea all how the resharpening processaffects the worm wheel surface.

A topographical chart call beobtained if a grid is made of pointsformed by profiles at different "face

(a IWorm su ria ce

. worm/gea r edges- u lines- 81ines

Fig. 5-Profiles lulines) and leads (ennes) of 81ZA worm wheel.

modification x.,amount

were measured at three different 0 values Fig.I6-AxlalsectioD ,ofZA..tyliB of WOmt("face posi.tiol1s"),and the leads weremeasured attltree different u values("diameters"). Different profile/leadtraces may have different numbers ofrecorded points when the same incrementof ufO' is used for the measurementsbecause of the shape of the wheel toothand the distortion of the mapping definedby Equation 13. As in the case of an invo-lute gear inspection, the profile/leadtraces on the same tooth. at different loca-tions should follow the same shape. IIIourinspection, it is found that the leadtraces are quite similar to each other,while the profile traces differ from oneanother. We believe this is due to cuttingscallops, These scallops create wavinesson a profile trace, and the waviness showsup with different phases at differentinspection positions. The effect of the

., " ..pIi"""""u{tMll

Ifg a i4p:liflllnwtu(fI"IM)

Fig. 7....Jprofiletraces of four wheel tletllat ~mid'die face position. ~

NOVEMBER/OECEMBER '199& 21

_-------------I:NSPECTIIONI------------_

Figl ,B-Lead traces of lou r whee'l teeth at "pitchdiameter," -

Fig. 9-ihrell profile traces of one tooth at threedifferent "face positions.,'"

Fig,. 10- Three lead traces of one tooth at threedifferent "diameters:22 GEAR TECHNOLOGY

scallops on the lead traces is not as largeas that on the profile traces partly becausethe scallops run along the lead direction.

Figure 11 shows the topographicalchart of the left flank of the first tooth. Intotal, 936 points were measured. Thepoints are the intersection points between40 leads (elines) and 30 profiles (slines).In Figure l Ia.the topographical deviationis plotted over a u x e grid, and in Figurellb, the topographical deviation is plot-ted on a R x F grid. (R is the radius of themeasured point, and F is its face posi-tion.) In Figure llb, the points with adeviation of -0.0254 mm are also drawn.The shape formed by these points can berelated to the contact pattern.

Measurement Reference FrameOne very important issue for the mea-

surement and inspection of a worm wheelis the establishment of the referenceframe. There are four types of referencemisalignments: eccentricity, wobble, the

,E 0.05E 0

'; -0.05.g 1...~-c

-3

-3.5

35 -4parameter e {radian)parameter ulrnm)

lal Worm wheel topographical chan lu X e grid)

points with -O.n254 mm deviationfrom the highe st point

radial distance (mm) Face position (mm)(a] Worm wheel topographical chart Ir X f gide)

fig. 11-Topog.raphical 'inspection of II wormwhee'! tooth: .

Fig. 12-8oHom land surface of revolution_

middle face datum and the orientation ofthe teeth. Each of these misalignmentsaffects the inspection results. As usual,the Z axis of the bore (or the shaft) is usedas the wheel axis. To define the middleface datum, one end face position can beused, and the distance from the middleface datum to this end face must be strict-ly controlled during hobbing. The orien-tation of the wheel can be defined by acenter direction of a tooth or the centerdirection of a tooth space. This can onlybe achieved by measuring points on twotooth flanks.

The effect of eccentricity and wobbleon profile and lead inspection is similar tothat of the inspection of an involute gear. Amiddle face datum misalignment intro-duces slopes on lead traces; the leads ofthe left side and the right side of one toothtilt in different directions. The wheel ori-entation misalignment introduces slopes ofprofile traces, and the slopes of the leftprofile and the right profile have differentdirections ..

To achieve a good middle face datum,the authors used the measurement data. ofthe bottom land surface of revolution (seeFig .. 12). The bottom land surface of rev-olution consists of all of the bottom landsbetween two consecutive teeth ..Its shapeis determined by three parameters: thebobbing center distance, the radius of thetop cylinder of the hob and the amount ofthe swivel angle. This surface of revolu-tion is symmetrical to the wheel's middleface position. If this surface is measured,surface fitting produces a very good mid-dle face datum. There are two restrictionsto apply this method: (1) the top blade ofthe hob must be straight; (2) the bottomland surface must be large enough forprobe access.

SununaryIn this article, we have presented new

definitions of profiles and leads of aworm wheel based on an explicit equa-tion of the generated wheel tooth surface.The new definitions are comparable totheir counterparts of an involute gear.Then, the inspection of profile, lead andtopography is discussed. Two methods ofinspecting profiles and leads are pro-posed, and an. inspection example of a

____ II.NSPECTIUNI _

ZK-type worm wheel is given. The proce-dure to define a good reference frame forworm wheel inspection is also discussed,The explicitequation of'the generated.wheel tooth surface derived i..11 this articleapplies to both right-axis and non-right-angle worm gearing. 0

AcknowledgmentThe authors would like to thank the

sponsors of the Gear Dynamics and GearNoise Research Laboratory at The OhioState University for their encouragementand financial support of the research dis-cussed in this article.

ReferencesL. Barber, C. "Wonn. Gear Measure-

ment." Gear Technology. Septem-ber/October 1997.

2. Hu, land J.A. Pennell. "A PracticalMethod of Measuring the WormWheels of Cylindrical Wonn Gears,"Proceedings of the 1994 Intern-ational Gearing Conference, I.Mech. E., New Castle, England.

3. Octrue, M, M. Denis and L. Faure."How to Inspect 'the Profile of aWO.nn Gear," Proceedings of the1984 ASME Design EngineeringTechnical Conference, Paper No.84-DET-204.

4. Litvin, F.L. and V. Kin. "Computer-ized Simulation of Meshing. BearingContact for Single Enveloping WormDrives." Journal of MechanicalDesign, Vol. 114, pp. 313-316.

5. Litvin, F.L. Gear Geometry andApplied Theory. PTR Prentice HaII,1994. pp. ]1-117.

6. SIl, X. and D'R, Houser, "CoordinateMeasurement and Reverse En-gineering of ZK Type WormGearing." Proceedings of 1997AGMA Fail Technical Meeting.Paper No. 97FrMS 1.

7. I7W Design Manual No.6, SpiroidGearing, 1986.

Ten Us What You Think ...If you found these Revolutions of interestand/or useful, please circle 201.

!Now You (an Meet ISO 90011Glealr lnspectionDemands With A Simple Mouse (Ikk

1

Yes, it's that easy to get accept/reject test results to ISO, AGMA and DIN standards with Mahr'snew DFI 890 series double flank gear roll testers, Easy-to-use, Windowsl '95"CompatibleWinGearx test and evaluation software let. you determine Total Composite, Tooth-to-Iooth andRadial Runout errors with a single mouse click.

There's much more, of course, including Mahr's field-proven modular mechanical design and,on the model 896, a highly sensitive leaf spring transmission which allows measuring forceadjustments to 0 ounces - an especially critical feature for testing plastic and powdered metal gear.

FrJl'FREE FACTS, ClJ'IlIact the Mahr gear measurement specialists:1-800-969'1331 Fax: 5131489,2020.

~The Measure of Your Quality

Mahr Corporation11435 Williamson Road : Cincinnati, OH 45241 Phone: 1800-9691331' Fax: 513/4892U20

Grll~ GerlT Wilh MaJ.,'s Metro/DID PrD,,~mFacloryjJrollm, fu1nd-ht/dg,., mtas.ring 1o.1s,' PC controlltd. doubkfomkroU Itslfrl' eNe alUl/ytitalgtar and

form t,Hm S.ifau finish wI t~u;pm!1!I jo, 8,r 100Mprofiln

CIRCLE 137

The Griell Geardini_ alles: IBurrs Vanishl Before Your Ibes'!

Of course, there's not really a magical way to handle yourdeburring needs, but we have the next best thing. OLSbuildsturnkey systems with proven performance, We've become theindustry leader by offering the best overall value:, High-speed, highquality systems,. Quality components from brand-

name manufacturers A variety of standard base models

which can be adapted easily toyour needs

Engineered solutions for practicallyany application

Trained staff 01experts waiting toassist you

OLS Model 1200

IC8IIIIOLS todal. We'lllma.e vour Ideburringtrou'bles disappear!

..'- '~ ,..

~ - ~".L;.~

CIRCLE 197NOVEMBER/DECEMBER 1999 ,23,

Skive hobbing,

hard skiving or

skive shaping can

be' alternatives to

ge.ar grinding for

post heat treet-

ment finishing,

vorable part geometries, however, heat treatmentleads to substantial hardening distortion. Becauseof today's standards for high quality gears, sub-sequent hard finishing of the tooth flanks is ohenneeded (Fig. 1).

The dominant finishing process used to com-pensate for hardening distortion is currently geargrinding, which can be used to machine very hardsurfaces with great precision. Despite successfulsophistication of grinding technology, machiningwith a geometrically undefined cutting edgeremains a time-consuming process with corre-spondingly substantial: machine and personnelcosts. There is, therefore, a desire to substitutemachining operations with a geometricallydefined cutting edge for the present grindingprocess. Sophisticated tool materials and cuttingsnow permit the use of defined-edge processes likepeeling, skive hobbing, hard skiving and skiveshaping to finish hardened tooth flanks.

Apart from higher removal rates, defined-edge processes have the advantage of combiningsome soft and hard machining operations on thesame machine. This enables the manufacturer tosave the purchasing costs for a gear grindingmachine. Another positive factor is the low ener-gy consumption for defined-edge machining.Finally, a d.ry cut is often feasible, eliminating theuse of cooling lubricants with their high disposal:costs and environmental risks.

One drawback of hard finishing with a geo-metrically-defined. cutting edge is lower processreliability due to the possibility of sudden toolfailure resulting from breakage at the cutting edge.This is caused by the relatively low toughness ofthe carbide tool material, The disadvantage of"low process reliability" and the drawback of"essential minimum chip thickness"are closelylinked. If the chip thickness is too small, no chipis cut; the work material: is merely pushed aside,increasing friction and pressure on the cutting

Hard Gear Finishing witha Geometrically Defined

Cutting EdgeProf. IDr..-I'ng:. Fritz Klocke and D:ipl.-lng.Thomas Kiinner

IntroductionThe market demand for gear manufacturers to

transmit higher torques via smaller-sized gearunits inevitably leads to the use of case-hardenedgears with high manufacturing and surface quali-ty. In order to generate high part quality, there isan increasing trend towards the elimination of theprocess-induced distortion that occurs during heattreatment by means of subsequent hard finishing.

Intensive research activity into the hard fin-ishing of gear flanks has produced an alternativeto the widely used gear grinding process.Manufacturers now have the option of choosing aprocess that uses a geometrically defined cuttingedge. This article describes problems and trendsin hard machining with a defined cutting edge,presenting the skive hobbing, hard skiving andskive shaping processes. The key feature is theelucidation of individual: process kinematics, toolgeometries, tool materials and coating systems. Anumber of fields of application are also indicated.

Hard Finishing Tooth Flanks WithA Defined. Cutting Edge

Nowadays, most gears are case-hardenedafter roughing in order to enhance their wearresistance and load-carrying capacity. With unfa-

I soft g,earclLlttilng (hobbing, shaping, ....) II I

I shaving II 1

I heat treatment IU 1 I 1

tMII~1 "1IdMIg J I gr1ndlng Ir~.1Igear hoolng' I shave gr1ndlng I

~ ,.. .. po V V V V VI mounting I

Fig. I-s-Finishing external cylindrical gears.24 GEAR TECHNOLOGY

edge and causing early failure. The minimum clilip

thickness needed for .81 chip to form is also .81 draw-

back. in terms of the accuracy-to-sizetaat can beachieved as compared to grinding. Defined-edge

hard cutting has difficulty in achieving the kind ofmachining accuracies possible with grinding.Especially high surface-quality requirements can

only be met to a certain extent. becauseprocess-specific deviations in the generating cut are mir-rored on 'the surface of the part.

Suitable Tool Material'sMachining hardened ferrous materials

demands tool materials with strength properties

that match the special needs of hard cutting tech-nologies and which possess adequate mechanical

and thermal shock re istanee, e peciaHy in dis-continuous cutting operations, Oreal hardnessand edge Il1bility,low adhesion. high thermal sta-

bility. adequate toughness and a homogeneousfine-grained structure are often conflictingrequirements impo ed by hard finishing on a tool.material. The choice of tool material is alsoaffected by economic considerations (Ref. 6).

Micrograin Carbides

We/Co.-based micrograin carbides haverecently become an important factor in gear-mak-ing technology. Carbides are sintered materialsconsisting o.f a oft metallic binder pbase (cobalt),in which the carbides-in this case rung [en car-

bide-are embedded. Microgram carbides of thesame composition but with carbide grain sizesbelow ] um possess greater hardness and resis-tance to. compressive stress than conventional car-bides with a grain. size of roughly .1 to 3 um.Tungsten-carbide and cobalt-based carbides withaverage tungsten-carbide grain diameters :s; 0.5.1IDl are termed ultra-micrograin carbides.

Micrograin carbides are also characterized bytheir very high bending and tensile strength, sinceboth hardness and bending strength can be raisedwith a WC-crystal size below ] 1JDl. Modem man-ufacturing technologies also produce extremelyfine-grained! homogeneous microstructores. Aproduct of (l"ii!sk:i lid. consisting, for example. of94% WC and 6% Co, measured by mass, achievesa hitherto. un attained combination of hardness(2000 HV 30) with bending strength (4000N/mm2) (Fig. 2), which would have been eonsid-ered impossible even a few years ago. (Refs. 4. 5& 11).

The rise .in hardness as me tungsten-carbidegrain size decreases reduces abrasive wear, whiletile 50% Increase in bending strength, which ha

positive effects on edge staibHily. and its suitabili-ty for machining hardened materials with mini-

i:;B~.at1._...-.0'771

S O.llli 0,1:5~i 0,'0D 0,05~~ o~~~~~~~~~~

_

Fig. J-Too/.workpiece C01lflgurat!OI!in skive bobbing, Source: PjfluJer.

4 ~ltlngl.olthftank culling edge

cr. tip rake, angle

~

Sldve Hobbi.ng Process Characteristicsand ApplicatiollS

Prior to skive bobbing, the tooth gap must berough-machined to a tage at which the tip of theskivi'ng hob does not make contact, and only theflank cutting edges are involved in the machiningoperation. Otherwise there would be ,8 risk ofchipp.ing ..The tooth root caa be freed by roughingwith an increased tip factor or by pre-hobbingwith a protuberance. The centering of the skivinghob in the workpiece gap is also of great impor-tance for the finished result Because of the highforces encountered in machining hardened steel.the skive hobbing machine requires high staticand dynamic stiffness in addition to geometric andkinematic accuracy.

Skive bobbing is carried out at cutting speeds ofvc '" 30-] 10 mlmin for any helix angle of the work-piece and modules Illn = 1-40 mm, The cuttingspeed has to be reduced aslhe size of the partincrea es, owing to the ,longer contact lengths in themachining process and the higher resulting thermalsire . Axial feeds range from 1....:5 m.mIworkpiecerevolution; small feeds are selected to match highworkpiece accuracy requirements. Owing to thesmall chip thicknesses concerned, skive hobbingshould. be performed in climbing cutting, in orderto reduce me tres on the cutting edge . The kivehob can be hifted during the hobbing operarion todistribute wear evenly over the tool.

Skive hobbing is used as a finishing operationor as a roughing process to prepare for subsequentfinishing by eliminating hardness distortions andreducing grinding allowances. This processsequence is frequently employed for large-moduleworkpieces, The quality limits are determinedmainly by the characteristic feed markings andenveloping cut deviations. Subsequent gear honing ican remove these on small-module gears. Gears !which cannot be ground because of their geometrycan be skive hobbed. Batch sizes range from one-off to large-series production.

Hard SkivingHard skiving (Refs. 1. 2, 3, & 7) is a continuous

defmed-edge process using an interrupted cut.Although its process kinematics are based on a gen-erating spiral drive as in the case of skivehobbing.hard skiving cannot be carried out on a skive hob-bing machine but requires the purchase of its ownmachine, Thi is due to the changed tool-workpiececonfiguration as compared to the skive hobbingprinciple (Fig. 5).

Because of the generating spiral. drive, the skiv-ing gearand workpiece me :h on skewed axes ofrotation. A spiral motion is uperimpo ed on the

spur-loothed: skiving ted:v,. voLJtan 13,

~, : peripheral speed

Vc : cutting speed

1'0, : helIx angleI: :axis intersection angle

Fig. 5-Too/.woril:piecf!' cOIl/iguratiorl in' skive' hobbing. SOUFCI.: Pfa!l/~1'.

rotation-speed-dependent generating motion. Thisreslllts from the displacement of the tool. paraUeilothe workpiece axis, entailing an additional. simulta-neous relative rotation of the workpiece. As a resultof (his superimposed generating and spiral motion,the cutting edges of the skiving gear slide along thetooth flanks of the workpiece and skive the materi-al over the fun width of the gear tooth. The inclina-tion of the skiving path to the generatrix on thetooth flank creates a surface strucrure favorable tothe noise behavior of 'the gears.

The cutting speed for hard skiving is a productof the difference in the circumferential velocitiesof the skiving gear and the workpiece. Owing tothe generating spiral drive, thecomponent Qf slid-ing velocity perpendicular to the cutting edges ofthe tool. is approximately equal to the cuttingspeed (Fig. 5). It may therefore be stated for a purskiving gear that

ve '" VuO -1:a:nB2If the helix angle of the gear B2 ;;;;0 ,Ihe cut-

ting speed wiUbe v c = 0 m/min, This means thatspur-toothed workpieces cannot be machined withspur- toothed skiving gears. Spur-toothed tools,however, do have substantial advllntages as com-pared to helical tools . For this reason, bar-d skivingis currently carried out only with spur-toothedskiving gear and is confined 1.0 the machining ofhelscal gear ..

Hard skiving tools. take the form of an under-cut cylindrical gear, because the proces kinemat-ics determine a cylindrical envelope profile of tileskiving gear. They con ist 0:1' carbide rings withsl.ightly conical or even cylindrical gear teeth ontheir external cylindrical surfaces.

Hard Sldving Process Charaeterlsttesand Applications

Hard skiving is carried out in a modale rangemn", 1-3 mm and at cutting speeds of Vc = 40-90mlmin and axial feeds 0.1-0.25 ram/workpiecerevelation. The machinable helix. angles of theworkpiece are restricted to 6.2'" 15-40". The 001l-

NOVEMaERIOECEMBER I~n 21

Fig. "i-The skiving tool can be reground in thee machine.Source: Pfauter

Fig. 7-Shapi/!g kinematics.

skive shapingnegativ tip ... ko, engle

shapinglpooitiv lip noke ."".,

..I .,..,.. ------------,.---'.1

'1'-.....,t=~L-+.r....,L,L:.::.:.::';:....~..lworkpiec:. ",n..-aIi!1g cul..-

Fig. 8~Negative rake angle Of! a cutting gear for skive shaping.ventional cutting is u ed for hard skiving. Anallowance ofO. 13 mm per flank can be removed inone cut. The allowance should be as small as pos-sible" in order to' save tool costs and keep processforces low.

Hard skiving tools are generally part-specific.A specially designed skiving gear i.s needed. foreach. workpiece ..The rake faces lie in a plane per-pendicular to the tool axis. This means that theskiving gear can be reconditioned at the end of itstool life by a simple, low-cost flat grinding opera-tion in the machine (Fi.g. 6).

Hard skiving is currently performed as a sin-gle-flank operation,even though two-flankmachining would allow shorter production times.One advantage of this approach is that the rightand left, flanks of the gear can be machined withdifferent shaft angles, prolonging the tool life ofthe skiving gear through larger effective toolorthogonal clearances: another is that workpiecequality can be improved bya more uniform pas-sive force curve over the axial path.

2B GEAR TiECHNOLOGV

Hard skiving is suitable only for medium- tolarge-scale production, especially for large-seriesproduction in the automotive industry. The gearsinvolved usually have profile bearing and crown-ing. Desired depth crownings can be taken intoaccount through a modified skiving gear profile.Crowning is generated by means of adaptedmachine motions.

Skive SbapingSkive shaping (Refs. 4, 9 & ] l) is a di contin-

uous defined-edge process that uses a translatorycutting motion to finish rough-cut. hardened cylin-drical gears. Lts process kinematics are identicalwith those of shaping (Fig. 7), enabling softmachining andhard finishing to take place on thesame machine.

The sequence of motions is characterized bythe generation between a cutting gear and work-piece mounted parallel to one another. Chips areremoved in the direction of the flank line throughan axial cutti:ng motion of the tool gear, referred toas a working stroke. During the subsequentreverse troke.the work table or the tool is liftedto prevent a collision between the cutting gear andthe workpiece. Skive shaping is performed with-out any axial offset, i.e. without lateral displace-ment. of the machine stand. This is the only meansof ensuring uniform infeed on both flanks .Collision problems that are encountered in rough-ing operations with a shaping machine do notOCCUrin skive shaping. owing to its nature as a fin-ishing cut (Ref. 9).

In order to withstand the initial cutting stress-es o.n hardened steel more effectively, the rakeangle of the shaping gear used in skive shaping isnegative. causing a negative tilt angle on the toothflanks (Fig. 8).

The involute tooth flank is profiled by meansof generated cuts, A cutting-gear tooth generates aworkpiece gap. The flanks of lite cutting-geartooth are involute in form and, as is in skive hob-bing, the tools are not part-specific except for nec-essary profile modifications. The face of a wornshaping gear is reground on a separate machine inorder to sharpen the cutting edge,

T0' achieve a generating motion, me workpieceand cutting gear move simultaneously with theaxial stroke motion, in accordance with the ratiobetween their numbers of teeth. The generatingfeed is the distance described at the pitch circleper double stroke (double stroke DS ". workingstroke + reverse stroke). The number of generatedcuts is dependent on the generating feed. Anincrease in generating feed reduces the number ofgenerated cuts.

JI

Skive Shaping PJIOCess'Characteristicsand. Applications

In . Idve shaping operations, it is necessary toensure that machining takes place only with theflank. cutting edges. The rol.lgh-cut workpiece gapmu t be machined appropriately and the euuinggear properly eentered, In order to prevent the tipcomers of the sliapiQg gear from participating in thecut. an allowance of 0.1 mmlflank: benld also beleft, Cul:l:in,g speeds are in the range vc= 2(}..4()mlmin. Gear deviations characterized by a piteh-cir-de gap are still an unsatisfactory feature. Profile-corrected tools and higher sliffnesse .of the overall.y tern may lead to improved profile accuracy.

Hard gear fillislling LDcases where the contactpoint is everely restricted by neighboring de ignelements may be very difficult or impo sible torealize. The special advantage of skive shaping.manifested in. the short tool ron-on. are apparentin the e application . Both contincus double heli-call teeth and double helical gears can be hard fin-i hed u ing this proee as can gears on teppedhafts or even crown gears. The machining of

chnch gears is also conceivable. Batch sizes maybe small or large.

One poten:l:iali application for skive bapinglie . in hard finishing of internal gear . Internalgears are a major component of planetary gearystem which represent an increasingly large

proportion of mass-produced gear systems.Owing 10 eobaacedperformance requirements forpower gear trains. users are demanding cost-effective hard finishing technologies for internalgears. Whereas hard broach:ing is economicallyfeasible only in large-series production, all unsat-isfactory tool transition, e.g. via neighboringde ign elements, is often the only ob tacle 10 theuse of ueh technologies as kive hobbing, geargrinding and gear honing. Skive haping, by eon-trast . i inherently : uitable for fini hing opera-tions on internal. gears.

C'onclusionsThe finishing of hardened tooth flanks after

heat treatment is of growing importance due toincreasing demands for .higb gear qUality andsmooth ronning. Currently. the dominant fini b-ing proce for elimination of hardening distor-tien i gear grinding. Sopbi ticased toolmaterialhowever, already point the way to defined-edgetechnologies.

Skive hobbing, hard skiving and skive shapingare defined-edge processes suited to the hard fin-ishjng of gears. These generating techniquemake use of micrograin carbide tools, Skive hob-bing tool are pirally hoped; hard skiving and

skive shaping tools are cylindrical. There are alsodifferences in process kinematic and potentialapplications.

Shve hobbing caa be used with any batch sizefrom one-off to large series production across abroad range .of geometrical workpiece dimen-ions. Hard! skiving. by comparison. is uited only

to series and large-cale production of heli a1gears with dimen ions cu tomary in the car indus-try. A common feature of both. processes is theirlow-noise surface structure.

In cases where the contact po.int i everelyrestricted by neighboring de. ign elements. hardgear finishing caeprefereatiatly be achieved bykive shaping. One potential application of kivehaping is the production of internal gears. which

have come to represent an ever-increa lng proper-tioo of gears manufactured in series productionand whic.h frequently cannot be machined byother processes. Skive shaping is fUlldamentallysuited to this task. 0

A,dcnowledgem nlThe authors grearfully acknowledge the "WZL-gcl!f

research circle" for the financial suppon of the studydescribed in this article.

References:I. Faulstich, 1 praUler - Zahruadhanbearbeilung mit

definierter Schneide, Pfauter VerzahDung se mi nar,Lud-wigsburg, 1996.

2. Faulstich, 1. Der feine Unterschied, N -Femguf\g. 1/88.8,74-81.

3. Faul tich, I. SchUlwtilzfrllsetl. und Hartschalen . Verfah-rensmerkmale und Einsatzfelder, Seminar "Pelnbear-beitung von Zahn.rlid.em-hart oder weich?". 16-17'September 1998. WZL der RWTH acben

4, Klocke. F, & 1. Kllllner. Ei.1lsatzpolentialebeimSchlllWlilzstol!en, Seminar "feinbea:rbeinmg von Z!!hn-riidem - hart oder welch?", 16-17 SeplC:mber 1998.WZL der R'WTH A chen,

5. Kolaska, H. & p, ElllTUlyer.Harrmeialle der neuen Gen-eration, METALL. 47. Jahrgang. Heft 10, October 1993.

6, Kolaska ..H. & H, Grewe, Pulvermetallarige derHartmetalle, Vorle ungsunterlagen, Hrsg, PacbverbendPulvermetallurgie, Hagen. 1994.

7. KOnig,.W. &F. Klocke. Fertigungsverfahren B!I!KIl, Ore-hen, frlisen, Bohren. VDI-Verlag GmbH. DO seldorl'l997.

g, loy. W. E. Hard GeM Processing with. Skiving Hob Gear Technology, MarchlApril1985.

9. Peiffer. K. WtllzstoBen einsarzgehdrteter Zyllnderrl1der,Dissertation RWTH Aachen 1991.

10. Roo V. Schlilwlilzfrlisen als Feinbearbeitungwerfnhreneinseugeharterer Zylinderrlider. Dissert lion RWTHAachen 1983.

I L Villlers. M. Hartfeinbearbeitung von einsarzgehllrlelenerzahnungen mit beschichteten HartmeUIIlen,

Dissertation RVIITH Aachen 1998.12. Young ..F. Carbide Rehobbi ng-A. New Technology Thai

Works!, Gear Technology. Mayllune 1994,

Till Us WIllI V. DIIIIc , ..If you found this article of interest and/or useful,please circle 2IIZ.

_ -INIDUS;TRVNEWS------------_Blankenship has a strong back- .

ground in both gear manufacturing andmetrology systems. Prior to joiningM&M. Blankenship served ]2 year with iGeneral Motors in a variety of technical !and leadership roles. most recently as iDirector, GM Gear Center and engineer-ing group manager for AdvancedPowertrain Development. He holds a doc-torate in mechanical engineering fromOhio State Uiliversity.

M&M P,RECIS'ION SYS:n:MS, N'AMESNEW IGEINERAILIMANAGER

i M&M Precision Sys-tems Corporation ofDayton. Ohio. hasappointed Dr. WeBlankenship generalmanager of M&M'sDayton

M

Model I(~H32-19LSlong Shah Gear Hobber

$79;99532" Diameter

19''' IFacel Width(Unlimited Length)

Model GS20-4TGear Shaper.$86.3951.2.1f' !Diamet'er4," (orB") face Width

Model HS10-12 CNC"eNe" Hob Sharpener

$,11, 2,~99510''' Diameter

12" !Length(ICBNI Wheell

CIRCLE 125

32 GEA.R TECHNO~OGV

____ IITECHNIICAIL CAlEN.DAR _

Novem.ber 8-12. AGMA Training School for 'GearManufacturing: Basic Course. Richard .L Daley College,Chicago, Il., The Training School for Gear Manufacturing pro-vides five days of Classroom and hands-on training. The diversecurriculum includes basic gearing. efficient machine set-uptechniques, accurate gear inspection and gearing calculation.For more information, can AGMA at (703) 684-0211 or visittheir Web site at www.agma.org.

November 9-12. Machine Parts '99. Shanghai. China. Aprofessional showcase exclusively for the growing machineparts industry in China, the show will have a full array of prod-ucts on display including standard and nonstandard fasteners,gears, chains, springs and powder metallurgical products, aswell as related processing equipment, materials and instru-ments. For more information. contact Busine s and Trade Fairs.Ltd. at (852) 2865-2633 or via e-mail atenquiry@bitj.com.hk.

November 14-19. 1999 International MechanicalEngjneering Congress and Exposition. Opryland HotelConvention Center, Nashville. T -. The Congress will includehundreds of technical sessions and exhibitors, Manufacturers,designers and suppliers will present products, services and tech-nology that will enable engineers and industry leaders to facethe increasing challenges ofthe next millenium, For more infor-mation contact June Leach-Barnaby at ASME at (212) 591-7795 or call toll-free at (800) 843-2763.

November 15--18. Gleason [Jfanter Burth Basic GearSchoo~. Loves Park, IL. A comprehensive, four-day course con-sisting of a coordinated series of lectures given byengineering production and inspection staff members. The course isdesigned for those new to gear manufacturing who are seekinga basic understanding of gear geometry, nomenclature, manu-facturing and inspection. Also held December 6-c9. For moreinformation contact Gleason Pfauter Hurth at (815) 877-8900.

December 1-3. Fundamentals .of Gear Design I. Centerfor Continuing Engineering Education, the University ofWisconsin at Milwaukee, Milwaukee, WI. This is Pan J of a 2-part course. covering the history of gear design, basic gear toothnomenclature, types of gears and their arrangements. the theoryof gear tooth action and failure modes and prevention. TheCOursehas been updated to provide a more comprehensive cov-erage of the important topics regarding the fundamentals of geardesign technology. For additional Information about this course,contact Richard G. Albers. program director, at (414) 227-3125.To sign up, contact the registration office at (414) 227-3139 or(888) 545-4700.

February 29-March 2, 2000. South-Tee Chadotte2000Advanced Productivity Exposition (APEX). CharlotteConvention Center. Charlotte, NC. SHIed as the South's largest.manufacturing event. For more information contact SME at(800) 733-4763 or visit their Web site at www..sme.org.

Tell Us Mit YouThink ...If you found this article of interest and/or useful, please circle 211.

http://www.agma.org.mailto:atenquiry@bitj.com.hk.http://www..sme.org.

we want to, work with! yOU!

CIRCtE1:n

The field-tested, PC user-friendly RC-400, hasquality, accuracy, small footprtnt, 3600 rotationalprobe, and is affordable .. Basic package inspects space, lead and involute,Many software options are available as well as cus-tom-designed packages.

Free Vid 0 AvailableRoto Technology, Inc.

35J F'IUllC Road, Dayton. OH 45449-2388TEL: (937) 859-8503. FAX: (937) 865-0656

www.rototech.com

-

ADVERTIISER INDEXFor more information about a product or service advertised in this issue of Gear Technology, circle the appropriate number on theReader Response Card and put the card in the mail.

For even faster response, use RAPID READER RESPONSE. Visit www.geartechnology.com/rrr.htm and have your requests for informa-tion delivered to advertisers instantly via e-mail!

-

ADVERTISER RS # PA'GE # ADVERTISER RS # PAGE It

III1t7169155124192144125156146

147, 112188148189

158.173IOrJ338337l30110

1IJ5. 157, 174332U6339333

112,176129161149

119,1771941.71160136

AIW System,Aju Magnelhel'f!HCAm arille Gear CompanyAmerican Metal TreatingAmerican Wero,ATA Gean.Barit, InternationalBasic lnccrporaied 'GroupBecket OearmeinersBerg. WM.Boom & Kocb Machine Tool Co.OI,p tan AtlanticCrown Gear B. V.Dr. Kaiser (S.L. hmsoolDura-BarErnugeEngi!1Cered Heal 'irealfls;;]erFellows CorporationGleason !'fauteI HunhGleason Pfauter Hur1h Cutting TbolsGraziano TrasmissioniGr6b_EmstHalifax Rack &. ScrewHermes AbraslvesHlmerHolroydlnductohearInsco Corpotl!tionITW HeartlandKappGmbHKoro Shnrpening ServiceKreiter GeartechleCount. Inc,

2063568685S68]3286M

8:1.47506719

45.494

497752

!Fe-!0]3.C.86,47

49145642

10,47751665

13.5211-9498653

M&M Precision SystemsMachine Too] BuildersMscsteelMahr Corporation,Micro GearMidwest Gear &. ToolMoore Measurement SolutionsMunson. S.L.National Broach &. MachineNiagara OeMNye LubricantsOn-Line ServicesP.H ..PrecisionParker IndustriesPC EnterprisePerry Technology Corp,Presrite CorporationProcess Gear Co.Pro-Gear CompanyThe Purdy CcrporanonQUlili,y TIO!II'mjssiOll ComponentsWyneRoto- TechnologyRussell Holbrook &. HendersonSpeedgrip ChuckStar Cutler Co.SUAmericaSuda International Gear WorksSurface CombustionTool inkUnique Power ProductsUnited Tool SupplyWabash MPI

164.165335104137336

15.1.178106.179

189186,187.193

1.62141:197191IS2163

134.181108IS3166

122, 182195

139.334131142

127.198128. 196. 342

107.183]67

115.184199341

170.185154

48.51695S2380

71.566,42

195.41. 57

86842J816787

mC,57316186

30,5243

15..433353

52.802.42.8758.43

8773,43

71117

85,4261

NOVEMBERIDECE'MBER 1999 33

http://www.rototech.comhttp://www.geartechnology.com/rrr.htm

Above: Spriral be-vel gearbeing ins,pectedl on a C'MM.

Courtesy of Brown & Sharpe.

IRight Worm being inspectedon a 'CMM. 'Counesv af

Brown & Sharpe.34 GEAR TECHNOLOGY

design and programming are strictly attunedtotesting gears. According to Mark Cowan. directorof metrology for M&M Precision Systems ofDayton, OH, "CNC generative gear testers areconsidered to be the fastest, most precise way 'Ofmeasuring parallel axis gears." Consideredturnkey systems, these machines are made so thatthe machinist on the shop floor can enter the nec-essary data and test the piece he is working on inaccordance with AGMA standards. Thesemachines are called generative because they workin much the same way as their mechanical prede-cessors, using a probe to p.hysicalJy trace out, orgenerate, the involute shape and lead. Accordingto Cowan, the process of CNC generative testingworks like this: "To measure an involute. theprobe is positioned at the base radius of the partand then driven along the linear tangential axis atthe same tangential velocity as the rotary at thatradius. So, you have a mechanical linkage thatdrives the probe along and actually generatestheprofile. The probe is actually scanning alongtbeinvolute, measuring deviation or error in the toothform .."

On the other hand, a coordinate measuringmachine records numerous axis positions as datapoints, which are then used to build-up the 3-Dmodel 'Of the part. "That's a more complicatedprocess." said Cowan. "because it's almo I. likemeasuring the part backwards. What the CMMdoes is move the probe around the tooth flank insome given plane. Theil it tits a theoretical formaround that plane and computes deviation fromthat For example, say you're moving a probe fromthe root diameter to the OD. If you're on a helicalpart, the actual place where the probe is contactingis moving because the helix angle on a gearchanges when you go from the root to the OD eventhough the lead stays the same. You have to com-pensate for that. Unless you have a canned programwritten by the CMM manufacturer, it's nota simplething to program a CMM t'Omeasure a gear."

Coordinate MeasuringMachines and the

Gear IndustryCharles ICooper

m ears are extremely complex shapes.Coordinate measuring machines, orCM.Ms. are designed to measure com-plex shapes. n seems to follow thatwould, therefore. be the ideal tool forCMMs

measuring gears. But the answer is not S'Osimple.While coordinate measuring may be the pre-

ferred way to quickly andaccurlltely inspect Ibevel gears, manufacturers of parallel axis gearshave long relied on their spline gages, roll testersand mechanical elemental. gear checkers for invo-lute and. lead inspection. More recently. they'veadded CNC generative gear testers and singleflank inspection machines to their repertoires.

But what about shops that make all types ofgears, or shops that need to measure splinedshafts, gearbox housings or 'Othercustom eompo-nents that a dedicated gear inspection systemcan't handle? F'Or these shops, measuring gears'Ona C~M might make sense, especially in lightof the recent improvements to their gear measur-ing software and their ease of use.

CNC Generative Gear Testiing Machines vs..Coordinate Measuring Machines

CNC generative gear testers are the most com-mon andpopular of the automatic, computer con-trolled machines used to perform analytic testingon gears today ..They are similar ill some respectsto coordinate measuring machines, but their

Because of this complexity. whi.ch off ets theflexibility that is the main strength of coordinatemeasurement machine technology. the machinesare con idered by many to be too, difficult to usefor regular shop floor personnel. According toRobert E. Smith. president of R.E. Smith & Co .Inc . Eli Rochester. NY-based gear metrology con-ultantand the co-chair of AGMA's Inspection and

Handbook Committee, "CMMs are very capablemachinesand can be very accurate. But to me, ittakes more of an engineer to run that machine thanII shop person. H you have a CNC generativemachine, most of the shop people would be capa-ble of running that." The rea on, according toSmith, is the complexity and user-hostility (asopposed to user-friendliness) ofjhe programming.and the time and skill it takes to etup and calibratethe equipment, "~ have lots of problems with the[CMM] gear software," said Smith. "For example,with some software you can't input DP. It's writtenfor module. So, if you're doing a gear here in theU.S., and it's the DP system, why should a guy sitdown with a hand ca.lcuJator when he's gOI this bigcomputer sitting there'?"

Of course, not everyone agrees with this per-ceived user-unfriendliness. "Ten years ago, you'dhave to sit down and write a Fortran program,"aid David Gene t, director of marketing and cor-

porate communications for Brown & Sharpe."Today it's much easier. You can download CADfiles or u e application-specific software." Ascompared to CNC generative te ters, 'Genest eesthe two technologiesand their d!riving software asfairly comparable, the only real difference beingthe familiarity of gear people with gear machines.

'CMM Ad'llan~gesCoordinate measurement machines have a

number of applications and advantage that gearmanufacturer should. understand and consider.These can be divided into two broad categories:bevel gears and qua.lity control.

Precision Bevel Gears. One area where coor-dinate measuring machine are needed is withbevel gears. According to Cowan, the main rea-on why CMMsare preferred for this work i that

bevel gears are not easily described in 3D. "Thenormal vector is constantly changing in threedimensions as you move from one place to thenext on the form of a bevel tooth. 'tau can't real-ly generate the motton, so you have to go to dif-ferent points to take it [the measurements]."Having to rely on data points means having torely on coordinate measuring machines to get thejob done. Smith agrees. "If you want 10 measurea bevel gear tooth shape, a coordinate measure-

HVBRIID !MACHINES-GEAR TIESTIERS OF TIHE FUTURE?Thera is an emerging class of machine that is neither a true coordinate

measuring machine nor a true eNC g,enerative g,eartester. These machinesare attempts to combine the' best features and abilities of both machinetypes in a single, integrated system that offers true, three-dimensional vol-umetric capabilities.

Next Dimension. Still underg,oing beta tests, the Next Dimension 430from Process Equipment Company is a dedicated gear tester that mixesCMM and generative taster technology to produce a very flexible gearmetrology tool. According to :DickConsidine, the software engineer on theND 4301project, "The ND 430' is a volumetrically mapped cccrcinate mea-suring machine that is designed to do gear inspection: The difference isthat this machine can measure features on a gear that traditional CNC gen-arative testers cannot in ways that other gieartesters cannot. "eNC gener-ative testers can give you relative measurements," said Considine. Wecan give you absolute, and relative measurements."

The software driving the machine is ,11 Windows 98-based package thatis entirely configurable to the customer's needs. ''With gear systems, oneof the things in the industry is that most software packages out therehaven't been designed. They have evolved. They've been around for a longtime and they've been added to until thareare now hodge-podges of stuffout there. Our system is brand new, the software totally rewritten and irsfully compliant with all three standards-ISO, AGMA and DIN."

Because it is so early in the machine's existence, the ,engineers atProcess are still figuringi out all the applications that the ND 430 will be ableto perform. "In addition to gear measurement," said Considine, "it will beable to do a lot of other things such as measuring cams relative to gears,keyways relative to g;earsand gears relative to gears on the same assem-blV.Normal geBr machines can't do that."

,Radiance 1006.The Radiance Radial Measuring System is designed pri-marily for rotary tasks. While this natura'lly includes gears, the Radiance isnot restricted to them. Rather, the machine is capable of scanning a varietyof cylindrical and rotary parts as well as hobs, shafts, cams and otherpieces. According to Jack Epstein, T8K product manager and the develop-er of the Radiance