B

Manutacturlnq of Forqed and Extruded

David J. Kuhlmann P.S.Raghupathi Battelle Columbus Division. Columbus, OH

Introduction ience and trial and error. A preliminary ment program is required for every new Traditional. methods of manufacturing die is made, and a few parts are formed, design, which makes the precision precision gears usually ,employ either Measurements are taken of the finished fanning process economically less at- or cutting. Both of these part, and the die is modified accord- tractive, especially when complex and processes rely upon generating the con- ingly, A second series of trials is con- precise geometries are involved, as with jugate tooth form by moving the work- ducted, and so on, until the final die gears. Therefore. methods need to be piece in a precise relation to the tool, geometry is obtained. Such a develop- developed to apply advancedeompu- Recently, attention has been ter-aided design and manufac- given to .forming teeth in a turing (CAD/CAM) single step. Advantages to such technologies (analysis of loads, a process include reduced pro- stresses, and temperatures us- duction time, material savings, ing finite element method based and improved performance approaches, etc.) to gear form- characteristics. Drawbacks in- ing die design and manufacture. elude complicated tool designs, This approach benefits hom the nan-uniformity of gears pro- capabilities of the computer in duced throughout the life ofthe performing complex mathe- tooling, and lengthy develop- matical analysis and mforma- ment times, tion storage, and allows the die Through projects funded by designer to examine the effects the U.S. Army Tank of various changes .0.£ process Automotive Command, Bat- variables on the die design, telle's Columbus Division without trying out each new developed a method for design- change on the shop floor, ing spiral bevel, spur, and helical gear forming dies. This CAD/CAM Applied To article discusses this method and and Extrusion summarizes the current state of En.!ml-. Fo;rm~na In recent years, CAD/CM1i the art regarding the manufac- , lood. lDAlStr ...... techniques have been. applied to I I!I'Id f-onn!"I [ntlty turing of farged and extruded various forging processes. The gears. experience gained in all of these applications implies. a certain Traditi.onal 'Gear overall methodology for CAD I .Fornting Methodology CANt of dies for precision and/ Design/test/modify. The or near-net shape forming. This traditional method of forging computerized approach is also and extrusion die design and applicable to precision cold and manufacture is based on exper- Fig. 1- Methodology for designing gear Forming dies. hot forming of spiral bevel, 36 Gear Technology spur, and helical gears, as seen inFig, 1, analysis of a metal forming process. For used to determine the friction shear fac- The procedure uses as input; the process a given microstructure, the flow stress, tors of various lubricants. This test in- variables and the part: (gear) geometry, a, is expressed as a function of strain, E, volves taking a ring of known dimen- The former consist of: strain rate, t and temperature, T: sions and temperature and upsett-ing it (1) data on billet material under (thickness reduction) between Hat dies, forming conditions (billet and die pro- a = f(e,tT) (1) Measurement of the final imide and perties, such as flow stress as a function outside diameters plus thickness enable of strain, strain-rate, temperatures, and To formulate the constitutive equa- one Itocalculate the friction shear factor heat transfer coefficients), tion (Equation 1) it is necessary to con- of the lubricant. Thus, to accurately (2) the friction coefficient to quantify duct torsion, plane-strain compression, predict the loads when forming gears, it the friction shear stress at the material and uniform axisymmetric-compres- is important that data be obtained and die interface, and sion tests. During anyone of these tests, regarding the friction associated with (3) forming conditions, such as tem- work creates a certain increase in the lubricant to be used. peratures, deformation rates, and sug- temperature of the billet material, which Estimation of Die Corrections, Tem~ gested number of forming operations. must be considered in evaluation and perature Effects. Temperatureaifects all Using the process variables and the use of the test results. in two ways. First, when the part geometry, a preliminary design of Tooling and Eguipment. Theselec- workpiece is heated to be able to form the finish forming die can be made. tion of a machine for a given process is the part, it expands .. In addition, the Next, stresses necessary to finish form influenced by the time, accuracy, and material will always increase in the part and temperatures in the mater- load / energy characteristics of that temperature due to deformation ial and the dies are calculated. The elas- machine. Optimum equipmentselec- heating. After the part is formed, it tic die deflections due to temperatures tion requires consideration of the entire shrinks during cooling, Thus, the final and stresses can be estimated and used forming system, including lot size, con- part size will.always be smaller than the to predict the small corrections neces- ditions at the plant, environmental ef- die in which it was formed. The second sary on the finish die geometry. Knowl- fects, and maintenance requirements, as temperature effect comes from the tool- edge of the forming stresses also allows well as the requirements of the specific ing. The 'tooling will always be slightly the prediction of forming load and en- part and process under consideration. heated due to the transfer of heat from ergy. The estimation of die geometry The tooling variables include design the workpiece. Inaddition, the die may corrections is necessary for obtaining and geometry, surface finish, stiffness be intentionally heated if the part is to close tolerance formed parts and for and mechanical and thermal properties be formed at elevated temperatures. the finish dies to exact di- under conditions of use. This prevents premature toe 1failure or mensions. The corrected finish die ge- Friction and Lubrication aithe T0011 defect formation in the final part. As at ometry is used to estimate the necessary Workpiece Interface. The mechanics of result, the part formed will always be volume, and the volume distribution in interface friction are very complex. One larger than the size of the die manufac- the billet or the preform. Ideally, a way of expressing friction quantita- tured at "room temperature". Both of simulation of the metal flow should be tively is through a friction coefficient, p" conducted for each die design. This is a or at friction shear factor. m. Thus, the AUTHOR: ------computerized prediction of metal flow frictional shear stress, 7, is ------~. DAVID J. KUHLMANN, is manager of at each instant during forming. This the Net Shape Manufacturing Section of Bat- simulation allows the determination of T = anop, (2) telle Columbus Division. For the past seven the cavity filling without excessive or years he has been developing interactiVe. loading of the dies and prediction of graphics- oriented computer programs for metal forming processes and luis authored defects. several papers. where an is the normal stress at the in- Process Variables terface, a is the flow stress of the de- DR. P.!;. RAGHUPATHI, former.lywith Billet Material Characterization. (l) forming material, and f is the friction Battelle Columbus Division, is an expert in For a given materialcomposition and factor. the area of cold extrusion, dosed die forging. deformation I heat treatment history Friction estimation. The friction be- deep drawing, metal forming machine tools, (microstructure), the flow stress, and tween the workpiece and the tooling is and computer aided design and manufactur- ing. In addition to being the author/co- in the workability (or formability) dependant upon the die surface, the author of more than 30 publications, he .is various directions (anisotropy), are the temperature.and the type of lubricant. also a co-editor of the Metal Forming Hand- most important material variables in A standard testcalled the ring test, is book (Springer- Verlag. 7985). July/Augustl990 37 these temperature effects must be com- • The flow stress, D', and the temper- The friction force, FE, is equal to the pensated for when designing tooling ature are constant within the sum of the friction forces shown in rig. which one expects to use to form near- analyzed portion of the deforming 2. The remaining forces, Fid and fs]"are net parts. material. computed from measurements of the Shrink fitting. Shrink fitnngrefers to The basic approach for using the slab flow stress of the material being formed. the tedmiqueof pladng the forming die, method is as follows. Once the total punch force is calculated, which contains the cavity in which the 1..The region of materia] undergoing the punch pressure, as well as the radial part is formed, under a residual com- plastic deformation is identified. pressure in the die cavity, can be deter- pressive stress to permit higher forming Thus, the boundary between the mined. This last parameter is required in loads. This compressive stress is elastic and plastic regions of the order to determine and compensate for generated by using an interference be- workpiece material is identified. die deflections. tween the die and one or more outer 2. One of the three principal stresses Case 2: Forging. A typical too] setup for rings. As a result of this shrink fitting, (assuming TRESCA's yield condi- forging spuror helical gears is given in the dimensions of the die cavity wiU be tion will be used) should be deter- Fig. 3 In this case, the slab method is smaller than when it was machined. mined. Inmost cases it will be easy applied in similar fashion to extrusion, This dimensional shrinkage must also to determine certain points where except that in this case, each tooth can be accounted for when designing the one of the principal stresses must be thought of as being extrudedindivid- forming dies for near-net shape manu- be zero. Por example, on at simple uaUy, and the total force calculated by facturing. forward extrusion process, the ax- multiplying by the number of teeth in E1asHc Deflection. (Mechanics of ial.stress should be zero at the exit the gear. The procedure for computa- Plastic Deformation I Slab Method). of the die. Itwill also be necessary tion of loads for forging spiral bevel During forming, pressure on the work- to determine the nature of this gears is also similar. (Fig. 4.) piece causes the die to expand, and the stress in the deformation zone material moves out to meet this ex- (tensile or compressive). Generating I:heGear Tooth W!ometry panded. surface. To compensate for this 3. In order to apply the TRESCA To define the tooth geometry, certain geometry change, the die must be cor- yield criterion, the maximum and gear and/or cutting tool parameters rected, depending on the estimated minimum principal stresses must must.be:spec...oedfied '.omeaUUltionS· ..J.J·,..·at u oarad-t pr,essure. To calculate this pressure, an be ascertained. pertaining to the mating gear may also elementary plasticity theory developed 4. The equilibrium equations of be required in certain instances. All the by Sachs(2)and SiebeP), known as slab stresses in the deformation zone theory, can be applied to practical metal are then formulated assuming the forming problems, The following rea- deformation zone is completely sonable assumptions are madeCll: filled .. Knowing the friction and , Punch 5~hoB<1 • The defomUng material is isotropic flow stress of the material. in the 2 Top I","", 67_Reirlforang . Ring - 3 Conical/Streamline

• The material flow criterion is Fp = F i.d + F", + Ff (5) defined by TRESCA's maximum where: shear stress criterion: Fp = total punch force, Fid = ideal deformation force, (4) FsIJ. = forte due to shearing of the material. at the entry andexit of where 0'1 and 0'3 are the maximum and the die, and fig . .2- Tooling arrangement for extruding parallel axis gears and the associated frlction I minimum principal stresses respec- Ff = friction force along the die walls forces. tively. and the punch. 38 Gear technology data required for the computations and Discharge Machining (EDM). The pro- passed through a straight wire that the geometry of the cutting tool can be cess uses an. electrode, usually made of moves in two dimensions, burning the obtained from a "summary sheet" graphite or brass, which is the negative die geometry as it moves .. Spur and developed by gear designers. With this of the die geometry, The electrode is straight bevel gear rues are the only ones data, standard gear equations are used brought dose to, but not in contact which can be cut using the wir-e EDM to calculate the X and Y coordinates of with, the die material. An electrical cur- process. In all cases, a corrected set of the points describing the gear tooth rent is allowed to arc across the gap gear tooth profile coordinates is needed. profile. which "burns" away the die material. Under contract from the United States Cutti.t)g the Die. A common method Another form of this method of manu- Army Tank-Automotive Command, of die manufacture is called Electrical facture is called wire EDM. Current is Battelle Columbus Division engineers developed two computer programs which calculate corrected gear tooth 1 Inner Punch 5 Support lnsert 9 Billel 2 Outer Punch 6 Stationa!)' Ring 10 Forging geometries and assist in :theoverall die 3 Top Insert 7 Reinlorcing Ring design pr-ocess. These programs, called. '* Helical Insert 8 Ejector SPBEVL and GeARDt are used for spiral bevel gear forging and spur/hel- ical gear extrustion Iforging respec- tively. Computer Program "GEARDl" ..The main functions of GEARDI are • Define the exact tooth form of a spur or helical gear, • Compute the forming load re- quired to produce the current gear design, • Compute the coordinates of the corrected gear geometry necessary for machini.ng LheEDM electrodes by taking into account the change

Before Forging Alter Forging in the die geometry due to temper- ature differentials, load stresses, Fig. J- Tooling arrangement for forging parallel axis gears. and shrink fitting, ..and, • Determine the specifieatiorts ofa tool which can cut the.altered tooth BEFORE FORGING AFTE.R FORGING 4 geometry on a conventional hob- bing or shaper cutter machine. This program enables the user to,de- sign spur and helical gears, predict tool- ing loads and pressures.estimate metal flow for forming the gear, and define the geometry required to manufacture the tooling using conventionalor wire EDM. Several examples of gears cur- rently being forged in industry were tested in the GEARDI computer pro- gram. The predicted forging loads were within acceptable limits of the actual t R.ngGear .2-6 Die Assembly loads measured during production 2 Die BonDm (W~h Teeth) 3 Inner Die Bottom '* Punch 5 Die Rir>g runs. 6 Doe Holder 7 Preform B Kick Out RII19 GEARDI is an interactive, graphics- oriented program which was developed fig ..4 - Tooling arrangement for forging spiral bevel gears. on a Digitial Equipment Corporation July/August1990 39 VAX 1 VMS 11/750 computer. It is a ticality of extruding parallel axis gears. formed in these trials is intended to be menu-driven program that allows the The spur gear was a 15 tooth, 5 DP an AGMA quality class B gear. Mea- user to 'select various options from a. gear. The gears were extruded using a surements taken on the extruded gears pre-defined list. The GEARDI program "push-through" concept. Each gear is indicated a gear of between AGMA has powerful application possibilities, first partially formed and left in the die quality 7 and 8. not only in the area of metal forming, while the punch is retracted. A second The helical gear chosen for the extru- but also in the area of gear and gear billet is placed on top of the partially sion trials was a 32 tooth 10 DP gear train design, with its ability to design formed gear and the press is cycled with a helix angle of 30°. During the hobs and shaper cutters and to modify again. During this cycle, the partially forming trials, several dies were clam- the fillet from a trochoidal shape to a formed gear is finished formed and aged from internal tooth breakage. This circular shape. pushed through the die, dropping out was due largely to the relatively high Computer Program ·'SPBEVI.:'..The the bottom of the die. Fig. 5 shows the helix angle and the fine pitch of the part .. main functions of the computer pro- sequence of parts in the tooling. Once Several die designs were used before gram "SPBEVL" are formed, the teeth on the gear are not parts were finally extruded sucoessfully. '. Determiae the tooth geometry for machined further. A fixture which Yet even these gears were observed to spiral bevel gears and pinions holds the gear on the pitch line of the have poor surface finish and were less normally machined on Gleason teeth is used to finish-machine the inside accurate than desired .. generators. The full irnplementa- and ends of the gear. The spur gear Spiral Bevel Gear. Refer to Fig. 4 for tion in the program is for FOR- a schematic of some initial forging tool- MA JElM gears only. This is ac- ing. The preform geometry is given in complished by an analysis of the Fig. 6..Hot work steel.H-13 was used as kinematics of the machine, blank the die material. Six electrodes were and cutter dimensions, and initial needed for the spark erosion of the die. settings proposed from the Gleason The gear forging trials were conducted summary charts. at Eaton Corporation's Forging Division • Correct the tooth geometry for the PARTIALLY in Marion, Ohio. A 3,(X)ij).ton,mechan- FORl'IED forgirng process parameters, such as GEAR ical forging press was used for the trials, temperature of forging, initial. die based upon a forging load estimate of temperatures, etc. The corrections 2,500 tons. The preforms were ma- are made for temperature dilieren- chined for these trials, though they tials, loading of the dies, die could be made by forming methods. manufacturing techniques (elec- FINISHED The preforms were heated in an induc- trode overbum, shrink-fitting,etc.) GEAR tion coil that was specifically designed • Give new machine settings, blank, for this purpose. The preform was and tool geometries determined heated for 200 seconds under a protec- geometry. tive atmosphere to the required forging frem the modified tooth Fig. 5 - Sequence of steps in extruding These are used to machine the parallel axis gears. temperature of 2,OOO°F(1,093°C). The EDM graphite electrodes on the forging loads were monitored using same Gleason machines that are used to cut the gears. 1------7.280 ------1 SPBEVL is an interactive gra- (184:912 i'nm) phics program operational on VAX/V1v1S systems.

EXtrusion Trials

Spur 1Hel..ica1 Gears.. Tooling de- 1.001 (25~ mm) 6.0lI0 -:- ---1 signed using 'theGEARDI program was (l54."32mm) I------~ 9.920 ------l used to conduct spur and helical gear (25196a """) extrusion trials at Battelle's Columbus I Division using a 700-ton hydraulic NOTE: RADII .375 UNLESS OTHERWISE NOTED press. These trials were intended to Preform Material: 8620 Steel validate the capabilities of the GEARDI Fig. 6 - Preform geometry for spiral bevel gear forgillg. program and demonstrate the prac- 40 Gear Technology For' pour .to,uglJest gear cutf-n:g Jobs,

t-llelJardest steel Is the easiest IclJolce' •.. CPM RBX 76

For hobs, shaper cutters and other ,gear ·cutting CPM REX 76 is just one of 10 high speed steels tools that are more than a cut above the rest, specify produced by the Crucible Particle' Metallurgy pro- Crucib'le CPM- iREX- 7,6. With 33% total alloy con- cess. With the industry's widest material selection, tent and an attainable hardness of HRC 68-70, this Crucible can meet your specific needs at any pro- h'igh speed steel prov,ides the highest avai:lable ductivity level, You can selectively upgrade to the combination of red hardness, wear resistance and best OPM material for the right application .. toughness, either coated or uncoated. On your next order, specilfy a high speed steel A 8ig,-Three US. auto maker recently realized a that's hard to beat . .. CPM REX 76 or another 300%· improvement in gear cuttingl toot Hfe by member of the CPM FlEXfamil.~To learn more, con- switching from M3 HSS- to CPM REX 76. With tact your nearest Crucible Service Center, or write' great,e.rtool life and excellent grindability. CPM REX Crucible Service Centers. 5639 West Genesee 76 means less downtime because resharpeninq is Street, Camillus, NY 130.31. easier and less frequent. Cruci:ble c:Service Center's, A Division of Crucible Maferlals Corporation

CIROlE A-1S ONI REAOER REPLY CARD BougLt & Sold: * BRIDLES * BELT * Bl'CKLES * BOLOS * CLOTHING * CHAP * BITS '* SPURS '* BOOKS '* SADDLES load transducers attached to the frames HIGH NO'ON of the press. W,est,ern CoJlectiLtes Die lubrica~ion, used during the forg~ (211) 202-9,01,0 by appoinunent onl» Lng trials, consisted primarily of a (2/3) 2,02·134,0 (fax) water-base graphite material sprayed with pressurized air ..After forging, the CIRCtE A-19 ON READER RiEPLYCARD gears wer placed teeth down in a sand- grapite mixture to reduce oxidation of the teeth during cooling. The back sur- faces were still left exposed to the air so that the cooling rates would not be ex- cessively slow. Gears were forged with a machining allowance of O.OO7~(0.178 mm) on both tooth surfaces. A specially designed nest located the teeth so that the back surfaces are machined to be reference surfaces. This reference sur- face was used in the finish machining of the teeth in a conventional gear cutting machine. The forged and machined gear was checked!ina Zeiss coordinate measuring machine ..Fig ..'J shows the relative devi- ation of the forged tooth profile as com- pared to a "master gear tooth" produced by conventional machining. Note that the relative error at the center of the pro- file is zero; i.e., the variations were mea- sured relative to 'the center of the coast and' drive surface of the master gear. The maximum variation was 0.003" (0.0762 mm), This difference can be trs true, our Consumer Information Catalog is tilled with booklets that easily compensated for in the machining can answer the questions American consumers ask. most. of the pinion. The above procedure was repeated to To satisfy every appetite, the Consumer Information Center puts together forge a 16~ ring gear. These trials this helpful Catalog quarterly containing more than 200 federal publications you can order. It's free, and so are almost half of the showed the applicability of the above booklets it lists. Subjects like nutrition, money management, health and approach in designing dies fOI" spiral federal benefits help you make the right choices and decisions. bevel gear forging.

So get a slice of American opportunity. Write today for your free Catalog: Conclusions & Future Work Parallel Axis Gearing. The GEARDI program has demonstrated the feasibil- ity of extruding and forging spur and helical gears ..Several U.S. gear manu- facturers are currently using this pro- 1 ,.~.Co... Department~.-u...m..er.. lo.ro.APf.mat..i.on.'.Center f: gram to design their dies on an ongoing 1 Pueblo, Colorado 81009 basis. Still, there is much to be deter- mined concerning the l:imitsof this pro- cess in terms of part geometry and pro-

"'public service of this pubhcanon and the Consumer Information Center cess conditions. Those companies which or the US General Services Adrmrusrraucn currently are forming gears using this 42 Gear Technolog:y CALENDAR C? .'It_ ."..... ~ .- ~.

SEPTEMBER 5-13, 1990. IMTS-90, McC'ormick Place, Ctd,eago,. It. Largest trade show in 'the Western hemisphere with exhibitors from around the world ..A new feature this year will be the 120,000 sq. fL Forming and Fabricating Pavilion .. For more inlormation, contact: [MTS ..90, 7901 Westpark Drive, McLean, VA 22102-4269. Ph:(703) 893-2900. Fax: (703) . 898-1151 . .' SEPTEMBER 11-1.3,1990. !I Short Course on Gear .1.'IIiNMENT MOOE - 0 Noise. The Ohio State

PROFILE MEASlJIUNG, RING GEAR OPERAT.: r .H ... ole University, Columbas, OH. Discussions will cover gear noise cause and fig. i-Zeiss gear tooth plots comparing the tooth Iorrn variation of a forged gear versus the cut reduction, measurement master gear. and analysis. transmission technology have gamed considerable sider the distortion of the gear tooth error methods. dynamic experience to date. There is a definite during cooling ..Tool life is another un- modeling, gear noise learning curve which 'one must travel in known factor. The fact that spiral bevel signal analysis, and shaft order to get into the business of forming gears are forged around the world (with and housing dynamics. gears. Research is still needed in many or without using the above method) is Case history workshops and laboratory demon- areas in order to generally quantify the an indication that the method suggested strations. Featured speak- process, including the following: in this article will be a good first step for ers include Profs. D. R. • gear materials which are formable, the initial design of the dies without Houser and R. Si.ngh of • lubrication, much trial and error. The future work OSU and Dr. Robert • tool life, is currently directed towards extending Munro of Huddersfield In- • variance in part dimensions be- the above concept to gears that are cut stitute of Technology, tween the first and last pieces from with both modified roll and tilt mecha- England. For more infor- mation contact: Mr. OJ given tool set, and nisms using Gleason machines. Richard D, Frasher, Direc- • limitations on part geometry. References: tor, Continuing Educa- The GEARDI program should also 1. AtTAN, T., OH, 5., and GEGEL.H. tion, College of Engineer- be improved to run on other computers, Metal Farming Fundamentals and Ap- ing, 2070 NeH Ave., Col- aswei11. asrmef iedt irutsi abil"lty tom odih_uy plications. ASM, Metals Park. Ohio. umbus, OH4!3210. (614) the fill.et Iormand to more accurately 1983, pp. 6-8,132. 292-8143. estimate forming loads and generate die 2. SACHS. G. "On the Theory of the OCTOBER 29-31, 1990. corrections. Drawing Process". (in German). Z. AGMA FaU T,echnical Bevel Gearing. One of the major angew. Math. tI. Mech., 7, 1927, pp. Meeting. Hilton Interna- 235-236. problems in spiral bevel gear forgi.ng is tional Hotel, Toronto. 3. SlEBEL. E. Metal Forming in Plastic the extension of the above approach to Canada, Papers and Condition (i.n German), Dusseldorf. seminars on gear design, other types of spiral bevel gears (gen- Verlag Stahleisen, 1932. analysis, manufacturing, erated tooth, [or example). Though the Acknowledgements: Reprinted with permis- applications, gea.r drives, modifications of the tooth surface can sioll of the American Gear Ma1ll4actl.!rers and related products. perhaps be determined by a rigorous Association. Tile opi71;oll5. silltemellts. and COI1- Contact AGMA. :1500 mathematical treatment of the problem, elusions presented in this paperar/i! those of tile King St., Suite 2101, Alex- authors and in no way represent the positiol1 or andria. VA, 22314. (703) the cost factor for making dies with an opinion of AGMA. electrode to be machined in a five-axis 684-0211. machine is still unknown. Further, the Special thanks to William L. /anninck for hls computer-aided methods should con- help witll tile technicalediting of this article. July/Augus11990 43,