Competition between Powder Metallurgy and Other Near Net Shape Processes: Case Studies in the Automotive and Aerospace Industries t
Narayan V. Nallicheri Project Manager, IBIS Associates, Inc.* Joel P. Clark POSCO Professor of Materials Engineering, MIT**
Introduction fabrication, and enhance material utilization With increasing industry competition, de rates, net shape and near net shape processes velopment of new alloys, stiffer price competi have been gaining in importance. These encom tion, and rising energy and material costs, there pass not only the traditional casting and forg has been a distinct trend to diverge from con ing, but also exotic powder techniques such as ventional forming technologies. The need to hot isostatic pressing, and powder forging. produce the final product in fewer processing One of the key forces driving the shift to steps with minimal material wastage has led wards near net shape production is the reduc engineers and component designers to develop tion of secondary machining or metal removal and critically evaluate alternative processing operations. The annual economic value of ma routes which would be more cost effective, terial removed, measured in terms of labor and while maintaining the same high level of per overhead, is estimated at over $125 billion in formance. Several techniques for manufacture 4 the U.S. alone ). Considering the enormous to "net shape" or "near net shape" are being economic value, it is apparent that any reduc developed to meet present and future demands. tion in metal cutting operations will provide The fundamental incentive for the develop substantial economic savings for the manufac ment of alternative fabrication processes is turing industry. To this end, there have been The reduction and, at times, elimi ~conomic. sizable efforts focused on improving the ma nation of a large number of machining steps chining operation and in developing means to provides sufficient cost savings on a per part reduce, if not eliminate, the amount of machin basis to warrant the use of net shape processes, ing required. Further, the field of machining even though they may entail extra capital has been aided by the development of a host of equipment and/or special handling equipment1). empirical relations relating the various cutting An indication of the significance of net shape parameters in order to choose the optimum processes can be found in automobile industry 5 6 cutting conditions • ) statistics: approximately fifty pounds of preci There have been a large number of changes sion formed parts are used in a typical Ameri in the field of machining itself, spurred by the can or European automobile, and about three 2 3 evolution of new materials, difficult to ma times that amount in a Japanese automobile • ). chine alloys, etc., and also by the need for The rna terial selection problem faced by pre higher precision. To meet these requirements sent day designers is also a problem of choosing there have been numerous advances in the cut the optimal processing route, especially since ting tool industry, such as the emergence of the choice of a particular material is tied to new material cutting systems and surface coat the manufacturing process employed. For ex ings aimed at enhancing tool life. Global com ample, a switch to steel from nodular iron for petition, in conjunction with the difficulty as connecting rod also implies a change in process sociated with machining some high-temperature ing from casting to forging. alloys and some composite materials, provides Driven by the need to reduce the costs of a constant incentive for the development of * 55 William St. Ste. 220, Wellesley, MA 02181, U.S.A. new tool materials and tool surface treatments ** Rm 8-409, 77 Mass. Ave., M.I.T.; Cambridge, MA 02139, aimed at extending metal removal rates. These U.S.A. advances serve to aid shifts in technology to t Received September 10, 1990. wards net shape. Figure 1 highlights the chron-
KONA No.8 (1990) 105 ological order of advances in the cutting speed tinual changes in engine designs place stringent capabilities of tool materials. The improve requirements on the possible materials which ments in cutting speeds progressively achieved may be used for their manufacture. The auto over the years have been both through the motive industry, where large production vol introduction of new tool materials (solid line), umes are typical, has been an optimal environ and the use of coatings on existing tool ma ment for the penetration of these technologies. terials (dashed line). The automotive industry is one of the largest users of formed parts, either cast, forged or stamped. In 1988, the automotive industry 10000 -----·------Polycrystalllne D1amond consumed a total of $16.9 billion worth of 0 5000 ------D S111con steel stampings, 1.6 million tons of nodular ~ Ceramic Nitride E 1000 ------CompOSite iron castings, and approximately 21 lbs/auto ? 0 Ceramic u 0 Coated Carbide of metal powder?). Thus, from the point of ~ 500------0 cubic~~~3~ view of introducing new net shape processing UJ Carbide OMicroqra1n (]) 1oo ------cast-AI-Io ____ ------;,;;;.-.:cr- carbide routes which may require high initial capital § ----- Coated HSS investments, it is preferable to have high pro (3 50------·H1gh Speed Steel---· Carbon Tool Steel duction volume runs to obtain economies of 0+------r----~------~-----, scale, which are typical of the automotive in 1800 1850 1900 1950 2000 Year dustry. In the case of P/M, almost 70% of the end of metal powder can be traced to the auto motive industry8>. Further, as mentioned above, Fig. 1 Chronological development of tools with the automotive industry is under continuous their cutting speed capability pressure to innovate and use lighter, stronger materials for its components. Figure 2 presents The technologies summarized above are just the trend of iron powder shipments used solely 9 a tip of the iceberg as far as the whole range of for P/M parts ). The increasing trend clearly possible fabrication routes available for metals shows the enhanced used of P /M parts in auto processing are concerned. There is an interde mobiles. In fact, General Motors recently an pendence between the technologies themselves, nounced their intention to use at least 18-20 to a limited extent, and between the primary lbs of powder parts per automobile in the next 10 fabrication routes and secondary routes, such few years ). as machining, to a larger extent. Although part The advent of new materials and near net of this shift towards near net shape technologies technologies is not limited to the automotive is material driven, a large fraction still stems industry alone. Table I tabulates the projected from the economic advantages to be reaped by growth rate for RS materials in the U.S. The a shift in technology. The advent of these new growth of these exotic materials is tied into the technologies warrants active markets where growth of the near net technologies associated 11 possible applications can be found. The range with their processing ). of applicability depends on the level of eco From the foregoing discussion it is apparent nomic savings offered by near net shape tech that there is currently a great deal of interest nologies, and on the performance characteris in the advent of near net shape technologies. tics obtainable. Further, a particular applica The prime motivation for a shift in technology tion may, at times, warrant the use of a high towards net shape is economic, though per performance material due to performance re formance considerations are also of importance. quirements. A change in material may also be The above discussion also suggests that there is associated with a change in processing route. a dual nature to the materials selection prob Thus competition between technologies may lem. Designers need to lay emphasis not only arise out of pure performance considerations on the level of performance delivered by the based on the limitations of the current material/ component, but also on the costs of manufac process. ture. The scenario outlined above is quite preva The drive towards more powerful, fuel effi lent in the automotive industry where con- cient aerospace engines has been such that,
106 KONA No.8 (1990) 200 ------ever, the next decade will show an increasing 180 ------"""' -- competitiveness among consortia as designers, 160------builders, and marketeers of engines.
KONA No.8 (1990) 107 largely limited by the presence of defects in will lead to an escalation of the amount of raw the alloy. The defects in a typical Rene 95 material required to process each component, alloy are determined by the processing route thereby leading to enhanced cost. undertaken. As shown in Table 2, the largest It is evident from the foregoing discussion defects in the as-hipped product are large re that while powder metallurgical processes, viz., active prior particle boundary defects. The in hot isostatic pressing and isothermal forging, clusion of small ceramic particles during the are being preferred over conventional wrought melting and atomization stage of the cycle techniques for this application, the exact limits the fatigue capability of the thermo choice depends not only on the economics of mechanically processed disk. As evidenced by fabrication, but also on the technical ramifica the data in Table 2, a shift in technology from tions of certain processing variables. This study hot isostatic pressing to isothermal forging was focused towards understanding the eco from extruded bar has potential for improve nomic and technical forces underlying the use ments in LCF properties. of these technologies. The automotive connect The mechanical properties of extruded com ing rod and the aerospace turbine disk were pacts were higher than their hipped counter Table 2 Comparison of defect type vs size at LCF parts due to the fine grain size of extrusions. initiation site Because of heavy deformation occurring in the HIP vs Extruded + iso-forge Rene 9 518 ) extrusion process, the size of the original pow Defect size, mils der particle did not significantly affect final Total 16 grain size ). Coarser particle size distribution Temp strain Defect HIP Extruded+ tF) range type* iso forge results in larger grained compacts having lower % tensile properties and increased stress rupture Avg. Max A vg. Max problems. Compaction by extrusion minimizes 1000 <0.75 1 6.7 36 4.8 9.8 differences caused by original powder particle 2 9.5 50 8.9 12.1 18.7 242 None None size distribution. It is this phenomenon that 3 4 6.3 10.3 None None leads to an enhancement of final disk proper ties. 1000 >0.75 1 11.2 50 4.8 8.1 2 13.1 Ill 7.1 16.0 The problems associated with the larger 3 19.3 83.6 None None scatter of low cycle fatigue properties in as-HIP 4 5.3 6.8 1.0 1.5 consolidated Rene 95 have increased the use of 750 0.75 1 7.5 9.1 4.1 8.7 isothermal forging techniques for final fabrica 2 8.6 13.1 5.2 10.2 tion. This, in tum, has encouraged shear de 3 6.4 6.4 None None formation methods such as extrusion for the 4 None None 0.5 0.7 ------~ ---· ---- forging stock in order to break up ceramic in * 1 - Ceramics, 2 - Ceramic clusters, 1 3 - Prior particle boundaries, 4 - Voids clusions and to reduce overall defect size 7). The large particle size variation for the hipped product has deleterious results on mechanical Table 3 Global materials consumption in aircraft 14 and fatigue properties. As opposed to this, iso gas turbine engines 1988-1998 ) thermal forging from an extruded bar results in Material Millions of pounds a more uniform grain structure, because the Superalloys 600.00 extrusion process breaks down the individual Steel 356.00 grains to a fine structure. This leads to an en Titanium 321.00 hancement of properties. Of course, it is possi Aluminum 75.00 ble to use a starting powder stock for hot iso Magnesium 10.00 static pressing with a very close size distribu Polymer/Polymer composites 13.00 Rubber/Elastomers 0.15 tion and cleanliness. This will result in en ------·------hanced properties of the isostatically pressed Total 1,375.15 turbine disk. However, a close powder particle size distribution will result in low yields at the used as cases to test the applicability of materi screening stage of the isostatic pressing opera als systems analysis at looking at material com tion, resulting in poor material utilization. This petitiveness.
108 KONA No.8 (1990) obtain a fine grain size (1-1 0 ,urn) stable at Powder Metallurgy of Superalloys temperatures above half the melting point, the The use of nickel base superalloys in engine powder compacts are extruded at temperatures disks has created widespread research interest, slightly below the recrystallization temperature especially in the area of processing. Powder of the alloys. metallurgy (P/M), has increasingly become the Hot isostatic pressing (HIP) is a technique process of choice, due to the flexibility it pro used to consolidate powders to full density. vides to tailor properties according to need. HIP is a process by which an isostatic pressure Further, P/M has been aided by advances in the is applied at high temperature, to metal powder field of rapid solidification, now capable of de placed within a container or die. The compact livering fine uniformly sized powder with is compressed in a pressure vessel using argon exotic non-equilibrium microstructures. The gas. Pressures range from 20 to 300 MPa, and mechanical and fatigue properties are often temperatures from 480°C for aluminum alloy tied closely to the size and size distribution of powders to approximately 1700°C for tungsten the raw powder used. This has led to research powder21 ). Processing temperatures are within into the production of fine powder. A con the plastic range of the material being formed trolled inert gas P /M process has proven suc - high enough for diffusion bonding to occur, cessful for producing fine grain, segregation yet low enough to prevent undesirable micro free material. Inert gas atomization requires the structural changes. Heat transfer during iso use of special controls, especially the control of static pressing is highly efficient. At high pres oxygen and nitrogen content which degrade sures argon gas is denser than water. To ensure mechanical properties if present in quantities uniform heating, molybdenum resistance heat larger than a few hundred ppm. Further, the ing elements are located throughout the work presence of certain tramp elements may be de zone22). trimental to processing, especially elements The first use of HIP was in the 1960's for such as Cr, Al, and Ti which form oxides that diffusion bonding of clad nuclear fuel elements. are difficult to reduce. Likewise, the nitrides of Consolidation of beryllium metal powder to Ti and Zr are detrimental as they do not dis shape followed shortly thereafter. Since then, sociate in a vacuum melting furnace. the size of units and use of the process has ex Conventional powder consolidation tech tended greatly. Uses now include the produc niques are not readily applicable to superalloy tion of titanium and superalloys for the aero powders because the powders are incompressi space industry, net shapes in PM beryllium and ble. Additionally, these powders do not sinter niobium alloys and other refractory metals. easily because they contain aluminum, chromi Isothermal forging and hot isostatic pressing um, and titanium, which oxidize easily at the have evolved into the dominant processes for sintering temperatures. Consequently, other this application over the past few years. An in techniques for powder consolidation which can vestigation geared towards understanding the apply high pressures at sintering temperatures, key forces underlying the choice of either of without triggering material problems, are re these technologies would be invaluable, due to quired19). the fact that they form a class of emerging near The superplastic behavior of superalloys was net shape processes. As such, a competitive discovered at Pratt & Whitney, and led to the analysis based on cost was carried out in this 20 development of the gatorizing process ). How study. The study used the technique of "Tech ever, it was found that gatorizing could not eli nical Cost Modeling" to estimate the manufac minate macrosegregation in as-forged alloys. turing costs based on the given fabrication Thus, gatorizing, more commonly known as routes. Performance also plays an important isothermal forging, was applied to dense PM role in the final choice of the processing route. preforms. On an industrial scale, it has been Therefore, the ramifications of performance applied to extruded PM billets. The super considerations on cost was also assessed. plastic deformation helps achieve forgings close Technical Cost Modeling to net shape, minimizing the amount of ma chining required and scrap losses. In order to The importance of a systematic evaluation
KONA No.8 (1990) 109 of the costs of manufacturing cannot be over Cobalt ...... 7.0-9.0% emphasized. The approach used in technical Iron ...... 0.50% Max cost modeling is to separate the different cost Tantalum ...... 0.20% Max elements and estimate each one separately. Columbium ...... 3.30-3.70% This applies basic engineering principles, the Zirconium ...... 0.03-0.07% physics of the manufacturing process, and clear Titanium ...... 2.30-2.70% ly defined and verifiable economic and ac Aluminum ...... 3.30-3.70% counting principles. The models themselves are Boron ...... 0.006-0.015% developed using a spreadsheet methodology. Tungsten ...... 3.30-3.70% The use of a spreadsheet methodology to simu Oxygen ...... 0.015% Max late the fabrication process is very helpful, be Nitrogen ...... 0.005% Max cause it provides a flexible environment for Hydrogen ...... 0.001% Max estimating the costs of manufacture using exotic Nickel ...... Remainder materials and processes that are currently not The powder required for the isostatic press- in production, and because it is easy to use and ing process is first screened to a-15 0 mesh size offers visible parametrizing of the model. was (the size assumed in this study). Based A detailed account of the various cost ele upon the requirements of the turbine disk, the ments are presented elsewhere. Interested read screened powder size distribution may need to 23 24 ers may review the following references )· ). be finer (around-270 mesh). The screening step of the operation is typified by low yields, espe Hot Isostatic Pressing cially if there are stringent controls on cleanli Based on the premise that isostatic pressing ness and the final powder size distribution. The and isothermal forging have emerged as viable screened powder is then blended to homoge alternatives for the turbine disk application, it nize different batches of atomized grades and stands to reason that a detailed construct aimed loaded into a clean, evacuated container. This at understanding the economic competitiveness stem also serves as a vent for outgassing. The of the above processes would be very helpful in container is usually made of mild steel. The making strategic decisions about the processes stem is joined to the container by gas tungsten in the future. This paper undertakes such an arc welding. evaluation using the technique mentioned The loaded container is outgassed by heat above. Further, the economic implication of ing under vacuum up to about 700°F. The technical changes is also highlighted in this containers are placed in the HIP unit. The iso paper. A competitive assessment of a typical static pressing was done at 2050°F, under a T700 turbine disk was carried out as a part of pressure of 30,000 psi for a minimum of 2 this study. hours. Under this pressure, the heated con The T700 Turboshaft engine design includes tainer collapses, compacting the metal powder. four cooling plates and two isostatically pressed At 2050°F the powder particles become plas P /M Rene 9 5 disks. This engine is characterized tic and coalesce to form a solid alloy. After by improved fuel consumption, extended oper pressing, the fully dense parts are heat treated ating life and simplified maintenance. As used as follows: in the U.S. Army Blackhawk helicopter, the 1. 2090°F, 1 hour, salt quench T700 engine has accumulated well over 200,000 2. Aged at 1600°F, 1 hour, air cooled 26 operating hours. The composition of Rene 9 5 3. Aged at 1200°F, 16 hours, air cooled ). used for turbine disk applications is presented Age hardening strengthens the material 25 below ): through precipitation of the gamma prime
Carbon ...... 0.04-0.09% (Ni 3 Al, Ti) and the heat treatment develops 27 Manganese ...... 0.15% Max resistance to low cycle fatigue ). Mechanical Silicon ...... 0.20% Max properties of Rene 95 are influenced by the Sulfur ...... 0.015%Max rate of cooling from the solutioning tempera Phosphorus ...... 0.015% Max. ture. If quenched too rapidly, there is a danger Chromium ...... 12.0-14.0% of cracking the disk. If quenched too slowly, it
110 KONA No.8 ( 1990) will not obtain the desired property level. It is Table 4 Breakdown of cost for hot isostatically for this reason that a salt bath solution and pressed turbine disk quench was used. By Factor A hot isostatic pressing model was con $/part Percent structed to estimate the manufacturing costs based on the sequence of operations shown in Raw material $1,045.00 36.28% Fig. 4. The set of assumptions used in the cost Process material $27.39 0.95% Labor $857.03 29.76% estimates is outlined below: Energy $12.70 0.44% HIP -Assumptions Scrap credit ($3.21) -0.11% Equipment $461.45 16.02% Finished Weight of Disk 4.5 lbs Tooling $214.05 7.43% Annual Production Volume 125 parts/year Aux. equipment $18.03 0.63% Raw Material Rene 95 Maintenance $69.22 2.40% Raw Material Cost $22/lb Taxes $9.23 0.32% Labor Cost $16/hr Insurance $6.92 0.24% Building $162.47 5.64% Labor Overhead 40% ------Labor Productivity 85% Total $2,880.29 100.00% Working Period 240 days/year ------By Process
$/part Percent
~ Contarner Loadrng Raw material $1,045.00 36.28% Screening $247.32 8.59% Blending $53.74 1.87% Container loading $205.22 7.12% Hipping $401.65 13.94% Heat treating $69.36 2.41% Tumblasting $3.49 0.12% Machining $436.78 15.16% Inspection $9.71 0.34% Testing $245.55 8.53% Fig. 4 Operation sequence for hot isostatic pressing Building $162.47 5.64% of turbine disks Total $2,880.29 100.00%
It was assumed that components other than turbine disk were made using the same equip ment used to fabricate the disk. In other words the isostatic pressing facility was run in a non IZZI Other [] Testrng dedicated fashion primarily due to the low pro E Machtnmg [[) Heat Treat duction volumes. The powder yield during the CJ HIP Q) C!l Conta1ner Loadtng screening operation is a very critical determi :;:; 1 Cl Blenomg nant of the total cost of the finished compo 0:: [i] Screentng Q) • Matertal nent. Figure 5 shows the breakdown of cost 0. u; 0 by processing step as estimated by the model, ll assuming a screening yield of 40%. Material cost is the most cost intensive factor in the total cost of the component. The high cost of con HI Ping (Screemng Y1e1d 40%) tainer loading is a consequence of the high labor content. The breakdown of cost by factor Fig. 5 Breakdown of cost by processing step for the and process is tabulated in Table 4. hot isostatically pressed turbine disk
Isothermal Forging is also called gatorizing (as patented by Pratt & The alternative to hot isostatic pressing is Whitney). The disks are forged from an ex isothermal forging of turbine disks. This process truded bar of Rene 9 5. The extrusion process
KONA No.8 (1990) 111 results in a more uniform, fine grain structure. below: The implications of this fine grain structure Bath 1. HCl + HF + HN02 will be explained later in this paper. Bath 2. HCl + HN03 +Ferric Solution The process involves heating stock to a tem It should be noted that the cost of disposing perature of about 2050°F and forging at a very the etchant is a significant fraction of the total controlled strain rate in dies heated to the same 28 cost of the etchant ). While the actual etchant temperature. The best available die material is is priced at $1.25/gallon, it costs $1.30/gallon TZM molybdenum, an alloy that is very strong to dispose of it. at the required forging temperature, but also The isothermal forging operation is accom one that oxidizes rapidly when exposed to air. plished using a pair of TZM Molybdenum dies. Consequently, the isothermal forging press re Under the assumed manufacturing scenario, the quires a vacuum chamber for the actual opera dies cost $175,000/set and are replaced after tion and sensitive temperature controls to en 600 forgings. sure that the material being forged remains at The heat treating and testing operations the temperature where it is superplastic. Other were assumed to be similar to those in the case system features unique to the isothermal forg of hot isostatic pressing. Each forging was as ing press include a special vacuum furnace to sumed to be individually tested for room tem bring the stock to forging temperature, transfer perature strength, high temperature strength, units for moving the forge mults from the and stress rupture properties. The final inspec furnace to the dies, and press controls that tion operations included Brinell hardness test allow the programming of slow cross-head ing, surface etching to reveal surface flaws, and movements. dimensional tolerance measurements. The sequence of processing steps is graphi cally shown in Fig. 6. The assumptions used in the model are listed below: Isothermal Forging- Assumptions B1llet Operat1ons 1---1 Coat BN f-. I Rough !so-Forge I Finished Weight of Disk 4.5 lbs Annual Production Volume 125 parts/year Coat BN 1- ._I__ Et_ch_~l- I Tumblast Raw Material Extruded Rene 95 Bar Raw Material Cost $31/lb Fm1sh !so-Forge j Rough Mach1ne Heat Treat 1- I -- I Labor Cost $16/hr Labor Overhead 40% Fm1sh Mach1n1ng 1-1 Mech. Testmg 1-1 Remove Test Pc Labor Productivity 85% Working Period 240 days/year The breakdown of cost by processing step, as estimated by the model, is graphically pre Fig. 6 Flowchart of operations for the case of sented in Fig. 7. Isothermal forging accounts the isothermally forged turbine disk for almost half of the total cost of the finished The billet operations mainly consist of cut component, because of the high cost of the ting and grinding the billet down to the required equipment and the TZM dies. Table 5 presents size. The boron nitride coating prior to iso a breakdown of total cost by factor and process. thermal forging serves both as a lubricant and Unlike the case of hot isostatic pressing, ma as a parting agent that prevents the adhesion of terial costs were a far smaller fraction of total the forged disk to the die. Isothermal forging is cost. This reduction is a consequence of the achieved in two stages, the initial forge yield material yield increase associated with the ing a rough disk shape. After the forging, the process. component is etched using an etchant called Competitive Analysis "Heppenstal. '' The purpose of the etchant is to reveal surface flaws. This is actually a two bath Figure 8 pres.ents a breakdown of cost by etching process. Their compositions are listed factor for the two competing processes. In the
112 KONA No.8 (1990) Table 5 Breakdown of cost for isothermally forged suiting in a tighter powder size distribution, turbine disk lead to very high material cost. Molybdenum By Factor TZM dies, being expensive and difficult to ma chine result in high tooling costs. The isother $/part Percent mal forging press, costing nearly $7,000,000, Raw material $589.00 14.05% results in a very high per piece equipment cost. Process material $11.86 0.28% Figure 9 presents a graphical representation Labor $705.67 16.83% of the sensitivity of cost to annual production Energy $9.80 0.23% volume. The HIP costs are presented for three Scrap credit ($1.24) -0.03% Equipment $1,127.45 26.89% different screening yields. The results show Tooling $1,356.84 32.36% economies of scale because the fixed costs are Aux. equipment $1.33 0.03% distributed over higher volumes leading to a Maintenance $169.12 4.03% smaller contribution on a per piece basis. The Taxes $22.55 0.54% spike in the curve at a volume of around 500 Insurance $16.91 0.40% Building $183.49 4.38% parts/year is due to the requirement of an addi tional set of tools. As can be seen, the break Total $4,192.78 100.00% even point shifts to higher production volumes ------By Process with increasing HIP screening yields. In other words, at low screening yields (say $/part Percent 10%) the hipped turbine disk is less expensive ------Raw material $589.00 14.05% at volumes below 50 parts/year. At all volumes Billet operations $133.76 3.19% above this, changing to isothermal forging is BN coating $11.51 0.27% economically advantageous. As opposed to Isothermal forging $2,033.89 48.51% this, if the yield were to increase to 15% at the Etching $23.86 0.57% Heat treating $86.88 2.07% Tum blasting $15.44 0.37% Machining $715.37 17.06% -:::- 4 ro Inspection $11.79 0.28% Q_ '-- Testing $387.78 9.25% fh Building $183,49 4.38% s8 3 (j) 2 0 Total $4,192.78 100.00% '!' Q_ w Q (;) 8 1
[]Other 1D Testmg HIPmg (40% Y1eld) HIPmg (15% Y1eld) Isothermal Forg1ng E:J Machtntng ...... 0 Heat Treat Fig. 8 Breakdown of cost by factor for the turbine ...... Isothermal Forge (j) ...... 0 0 0 ••••• disk '!' ...... (:3 Billet Operattons Q_ ...... • Mater~al ...... screening stage of the hot isostatic pressing operation, the break-even point would shift to 110 parts/year. Thus, while low screening yields result in a closer control in the powder size dis Isothermal Forging tribution leading to a concomitant enhance ment of fatigue and mechanical properties, it Fig. 7 Breakdown of cost by processing step for the isothermally forged turbine disk also results in highly elevated costs. The original motivation for resorting to a powder process case of hipping, the breakdown is presented for was to reduce final costs by minimizing materi two different screening yields. The close de al input. This is defeated because a coarse par pendence of final cost on the screening yield is ticle size distribution leads to poor LCF prop clearly shown. Low screening yields, while re- erties.
KONA No.8 (1990) 113 Figure 10 presents the variation of the cost to reduce the cost of fabrication. Referring to of the isostatically pressed turbine disk with Fig. 6, it can be seen that tooling is a sizable screening yield. The base case isothermal forg chunk of the final total cost of the component. ing cost refers to the cost as estimated by the Thus, any reduction in the cost of the TZM die isothermal forging model using the set of as set would aid in reducing the total fabrication sumptions listed above. It can be seen that at cost. screening yields below about 16%, isothermal Figure 11 presents a set of scenarios present forging presents cost benefits over hot isostatic ing the sensitivity of cost to production volume pressing. As screening yields increase, isostatic at different die set costs. The figure shows that pressing becomes progressively less expensive. a reduction in the cost of the TZM dies from However, there is an adverse effect on proper $175,000 to $25,000 results in a 30% reduc ties at high screening yields. This leads to the tion of total cost at volumes around 100 parts/ conclusion that although isostatic pressing does year. present a feasible alternative and has been used in the past, production with adequate perform 6.5 ------· ance is not a feasible solution, especially when isothermal forging presents an attractive cost effective alternative.
OJ (.) 6.5 ------(l) 0::: ~e; 6.0 ------HIP (10% y1eld) ------(l - ;;; "' 55 0 0 u ~50 HIP (15% y1eld) :'5 4.5 .- -.--- :.-.-.~---- :.·- ·.:------"x·------.----- Ql Annual Product1on Volume (parts/year) 0::: 4 0 Q) (l 3.5 Fig. 11 Variation of total cost with volume at different ;;; TZM die costs for the iso-forged turbine disk 8 30 25+-~~~~~~~~~~~==~~ 0 0 0 1 0 2 0.3 0.4 0.5 0 6 0 7 0 8 0 9 1 0 Case Study: Material Alternatives for Automo Annual Product1on Volume (000 parts/year) tive Connecting Rod Fig. 9 Variation of cost with production volume for Conventional forging has traditionally been the case of the turbine disk the processing route of choice for this applica tion. Of late, powder forging has emerged as a
11 ------key process for this application. Powder forg ing yields a high density component with prop ~ 10 ...... erties akin to those of conventional hot forging . "" 9 0 0 Although this technology emerged in the early s 8 ------
Ql seventies, it has not been exploited as com u ------Ql pletely as one would have expected. However, 0::: 6 ------Ql Ford Motor Company has been using powder (l ;;; forged connecting rods ever since I 98 7, and 0 4 u production recently crossed the I 0.0 million mark.
0.0 0.1 0.2 0.3 0.4 05 0.6 Further, in the case of conventional forging, Screenmg Y1eld ( x 10%) the rods and caps have been forged and ma chined separately. Powder forging, on the other Fig. 10 Variation of HIP cost with screening yield hand, undertakes a single piece forging opera tion. There seems to be no real technical reason Based on the premise that isothermal forging as to why the rods and caps cannot be forged as is the most feasible alternative for this applica a single piece by conventional forging as well. tion, it stands to reason that efforts be directed Based upon these considerations, the three
114 KONA No.8 (1990) processing routes for the manufacture of the 6.5 ------. ------automotive connecting rod and cap were evalu 6.0 ated: steel forging a one piece rod, forging the ~ 55 '- rod and cap separately, and powder forging a {f) 50 ~ 4 5 one piece rod. (l) [L Q; o 35 ------:;;:-:-Powder Forgmg ------· ~ 3.0 a; 0 2.5 2 P1ece Forge - ... -.-- .. ------Q 2.0 ----~ 00 05 10 15 20 25 30 35 40 45 50 Q) u Q) Annual Product1on Volume (000 000 parts/year) 0:: Fig. 13 Variation of cost with volume for the unmachined connecting rod & cap
o The breakdown clearly depicts the cost One Pc. Forg1ng Two Pc Forg1ng Powder Forging advantages of powder forged connecting Fig. 12 Breakdown of cost by factor for the rods over the competing processes, in spite connecting rod & cap of the high material costs. ° Forging the rod and cap as one single piece The costs of production and their breakdown offered a cost savings over a two piece by factor are presented graphically in Fig. 12. forging. There are several critical features to note: o The high cost of forging can be attributed not only to the amount of machining re quired, but also to the fact that expensive, Table 6 Breakup of cost by factor for the connecting dedicated equipment is required. rod & cap Fully machined cost The fabrication costs estimated by the dif ------ferent models for the four cases are presented Combined Separate Powder in Table 6. Powder forging is clearly less expen forging forging forging sive than the other alternatives. Combined forg Variable costs $2.49 $2.48 $2.10 ing offers cost savings over separate forging due Fixed costs $2.91 $3.02 $2.96 to lower fixed costs because of the elimination of a separate line to forge them separately. Total costs $5.40 $5.50 $5.06 Although these are point estimates, it is pos Funy machined cost sible to use the cost model to evaluate the sen sitivity of these costs to critical assumptions. Combined Separate Powder Sensitivity with respect to volume was evalu forging forging forging ated and is graphically presented in Figs. 13 and Processing cost $2.06 $2.16 $2.94 14. The sensitivity of the unmachined compo Machining cost $3.34 $3.34 $2.12 nent cost to production volume is of particular Total costs $5.40 $5.50 $5.06 importance when evaluating a high capital, near ------net shape alternative to a low capital, machin ing intensive process. In particular, the percep Combined Separate Powder forging forging forging tion of cost effectiveness can be adversely af ------fected if the consumer of a semi-finished part Material $0.89 $0.87 $0.96 is not familiar with machining differences be Labor $1.48 $1.49 $1.06 tween alternatives and their cost consequences. Tooling $1.31 $1.37 $1.57 Capital $1.21 $1.25 $1.06 For example, a process which may look cost ef Maintenance $0.24 $0.24 $0.21 fective to a consumer who buys less expensive Other $0.26 $0.27 $0.21 as-formed parts may actually entail enough in house machining that the final part cost ex Total $5.40 $5.50 $5.06 ceeds that which would be incurred using a
KONA No.8 (1990) 115 12 ...... made, it was virtually impossible to reverse the
'2 11 choice of material for the rod and cap, even ro 0. 10 though an alternative might offer some attrac ~ tive features. Further, increases in tensile and (]) 0 fatigue strength served only to improve the de £ 8 sign margin of safety in the engine. An interest ~ 7 ing feature was the fact that, although a stiffer 8 6 () rod may be beneficial up to a certain limit by 5 reducing vibration, further increases may actu 4+---r---r---.-~----~~---r--.---. ally be detrimental, since they can lead to in 0.5 1.0 1.5 2.0 2.5 3.0 3 5 4.0 4.5 5.0 creases in crank loading. Annual Production Volume (000.000 parts/year) Based upon the preceding cost analyses, the Fig. 14 Variation of cost with volume for the fully powder forged rod seems to be well positioned machined connecting rod & cap to substitute for conventionally forged rods. However, there is considerable skepticism as to more expensive, but nearer to net shape part. the cost effectiveness of these rods and caps. At the base case production volume of As was pointed out in the section on cost analy 3,000,000 parts/year, powder forging yields sis of connecting rods and caps, as-forged rods the most cost effective component. Figure 13 are cheaper than as-forged powder rods, while clearly shows that, in the unmachined state, in the finished state the reverse is true. At pre the powder forged rod and cap are more expen sent, the machinability of powder forged rods sive than the conventional rod and cap. This is an issue of some concern and there seems to is attributed to the high capital cost and the be a lack of comprehensive information in this cost per pound ofraw powder, which is almost area. Since it is the machining step which pro double the cost of bar stock. Further, it is also vides the cost benefit of powder forged rods, evident that forging the rod and cap as one uncertainty in this area is almost guaranteed to piece offers savings at all volumes considered slow introduction of this material technology. up to 5,000,000 parts/year over the other two Further, there is the issue of tooling expen cases. ditures on existing facilities to be considered As opposed to this, as seen in Fig. 14, pow when deciding a possible shift from conven der forging offers cost savings over convention tional forging to powder forging. In the event al forging at all volumes up to 5,000,000 parts/ of setting up of a new facility for connecting year. Powder forging offers a near net shape rods, it is quite apparent that opting for a pow product, unlike the conventionally forged rod der forging route as opposed to conventional and cap. This is clearly evidence by the cost forging offers substantial cost savings. savings accrued in the machining step of the The possibility of a combined forging as op operation. Thus, while powder forging was posed to a two piece forging was analyzed in more expensive that the conventionally forged the preceding sections. Stemming from the cost rod and cap in the unmachined state, it offers minimizing drive in decision making, it follows significant savings over the conventional rod that a shift to a one piece forging would im and cap in the fully machined state. This is a prove cost effectiveness. In fact, this practice is classic example of the cost savings accrued by being followed for some V-8 connecting rods a shift towards a near net shape technology. and caps. The cost difference between a two-piece and Cold forging of caps also offers the possibili one-piece conventional forging is about 9 cents/ ty of forging closer to net shape and better part, which is too small to detect within the forging detail, reducing cost by reducing the scale of the graph in Fig. 14 A primary factor number of machining steps. In situations where influencing materials choice in this application the rod and cap are forged separately, cold forg is the issue of tooling requirements for machin ing could improve cost effectiveness. ing of these rods. It became apparent that, This outlines the major issues of concern for once a commitment to tool a line for machin connecting rods and caps. While powder forg ing a rod and cap of a given material had been ing does not seem to be a potent threat to exist-
116 KONA No.8 (1990) ing facilities, it should be borne in mind that high in the unmachined state, due to the high the industry is becoming more aware of the cost of powder and capital equipment. How cost effectiveness of this processing route. Thus, ever, they offer significant cost savings over the for the case of green field facilities, it is likely conventional forged rod and cap in the fully that powder forging will be seriously considered, machined state, due to the fact that powder provided that both forging and machining be forging results in a near net shape component, conducted by the same source (since the major which does not require extensive machining. cost advantages come about only in the ma Thus, the savings are accrued in the machining chining stage of the operation). stage of the operation. While powder forging Discussions indicated that powder forging is has the highest degree of acceptability at the the only commercially acceptable alternative to moment due to its low cost, it is evident that forged connecting rods in higher performance redesign of the component resulting in weight engine applications. While more exotic material savings would further enhance this acceptable alternatives do exist and are in use, they are re rating due to cost savings. stricted to specialty platforms or long term de The case of the turbine disk presents an in velopment programs. teresting situation of competition between To restate the primary conclusions of the alternate net shape technologies. While isostatic utility analysis of connecting rods and caps: pressing can potentially provide low cost tur 1. The automotive industry is giving in bine disks, it requires high material yields in creased consideration to powder forged the screening stage of the operation, which in rods and caps. turn leads to wide range of particle size distri 2. The critical limitation to P/M competitive bution and degraded mechanical and fatigue ness in this application is the lack of com properties. On the other hand, isothermal forg prehensive, reliable process information, ing offers an alternative route to fabricate the especially machining information. component having acceptable performance. An 3. The economic advantages of powder forg isostatically pressed disk with performance ing can be best exploited by fabricators characteristics equivalent to that of an isother who both forge and machine the compo mally forged disk, requires low screening yields, nent. increasing cost dramatically. Thus, isothermal 4. Manufacturers who are currently tooled forging or gatorizing offers an optimal set of to handle conventionally forged rods and characteristics for commercial use in this mar caps can more cost effectively manufac ket. Efforts to reduce the cost of the isother ture connecting rods from one piece forg mally forged disk should be focused towards ings, rather than forging the rod and the reducing the cost of the TZM dies. The under cap separately. lying assumption in the above analysis is that 5. Steel forging shops must enhance forging the level of testing required is the same in both detail, reduce trim scrap, and move to cases. ward increased use of forging presses. Isothermal forging has a promising future if Better dimensional control and the reduc one could exploit the tremendous savings in tion of process scrap can go far to im material utilization possible. In the aerospace prove the economics of forging. industry, this accounts for a significant frac tion of the total cost of fabrication. Parts with Conclusions pronounced circular geometry are particularly This paper has highlighted the growing im well suited for isothermal forging due to the portance of near net shape processes in metals fact that the symmetry of the component aids fabrication. The economics of near net forming in prolonging the die life. Smaller size and low has been treated using the very useful tool of er complexity makes it easier to forge the sur cost modeling. faces to net shape. In large forgings over about The connecting rod and cap offers a classic 150 sq.in. plane area, net surface forging be example of the cost savings to be accrued from comes increasingly difficult because of the opting for a near net shape production route. manufacturing tolerances on die dimensions, The cost of powder forged rods and caps is variability in material properties, and varia-
KONA No.8 (1990) 117 29 tion and non-uniformity of die temperatures ): 15) Private Communication, Netcast, Investment Cast Research geared towards reducing the high cost ing Institute, January 1990. of TZM tooling or using alternate tool materi 16) Baker, J.F. and E.H. VanDerMolen, "Effects of als would also help enhance the economic posi Processing Variables on PfM Rene 95", Super tion of this process. alloys Processing, Proceedings of 2nd International Conference, September 18-20, 1972, Champion, Pennsylvania, pp. AA1-AA25. References 17) Gessinger, G. H.: "Mechanical Properties of dis I) McGee: S.W. and F.K. Burgess, "Identifying Cost persoid free P/M superalloys", Powder Metallurgy Effective P/M Applications", International Journal ofSuperalloys, Butterworths, 1984, pp. 153. of Powder Metallurgy and Powder Technology, 18) Ross, E.W., D.R. Chang, R.A. Sprague and P.J. Vol. 12, No.4. Linko: "Superalloys in 2001 ", Advanced High 2) Production to Near Net Shape, American Society Temperature Alloys - Processing & Properties, of Metals Source Book, Ed. C. J. Van Tyne and B. Nicholas J. Grant Symposium, Ed. S.A. Allen, Avitzur, American Society of Metals, Metals Park, R.G. Pelloux & R. Widmer, American Society of Ohio, 1983. Metals, 1986, pp. 152. 3) Van Dommelen, H.M.: "Net Shape Forming", Pri 19) Gessinger, G.H. and M.J. Bomford: "Powder vate Report. Metallurgy of Superalloys", International Metal 4) Koster, W.P.: "Machining-Forecast '90",Advanced lurgicalReview, Vol. 19,1974, pp. 51-77. Materials & Processes, Vol. 137, Issue I, January 20) Pratt & Whitney, U.S. Patent No. 3519503. 1990, pp. 67. 21) Gane, N.: "Developments in Powder Metallurgy", 5) Boothroyd, G.: Fundamentals ofMetal Oming and Materials Forum, Vol. 13, 1989, pp. 81-100. Machine Tools, McGraw Hill publications, 1975, 22) Hanes, H.D. and J.M. McFadden: "HIP'ping of pp. 142-163. Castings: An Update", Asia Pressure Systems 6) Machining Data Handbook, American Society of Monograph. Metals Publications, "Economics in Machining and 23) Nallicheri, N. V., J.P. Clark & F.R. Field: "A Tech Grinding", pp. 1-41. nical & Economic Analysis of Alternate Manufac 7) Predicasts: Basebook 1987, Predicasts, Issue No. turing Processes for the Connecting Rod", Pro 112, July 15, 1988. ceedings, International Conference on Powder 8) Industry News, International Journal of Powder Metallurgy, Pittsburgh, May 1990. Metallurgy, Vol. 22, No.3, 1986, pp. 138-139. 24) Bush, J. V. and F.R. Field: "Technical Cost Model 9) Johnson, P. K.: Metal Powder Industries Federa ing", Blow Molding Handbook, eds. Donald & tion, Private Communication, August 1989. Dominick V. Rosato, Hanser Publishers, 1989, 10) Metalworking News, March 5, 1990, pp. 1. pp. 839-870. 11) Froes, F.H.: "Powder Metallurgy", Advanced Ma 25) Happ, M. B. and P. S. Clemente, "HIP of Rene 95", terials & Processes, Vol. 137, Issue 1, January U.S. Army Manufacturing Technology Journal, 1990, pp. 56. Vol. 4, No.4, 1979, pp. 41-44. 12) Driver, D.: "Near Net Shape Manufacture of Aero 26) Eisen, B.: Crucible Compaction Metals, Private Engine Components", Metals & Materials, August Communication, January 1990. 1988, pp. 492-497. 27) Lasday, S.B.: "Hot Isostatic Pressing of Heat and 13) Driver, D~ "Developments in Aero Engine Materi Alloy Corrosion REsistant P/M Alloy Components als", Metals & Materials, June 1985, pp. 350-352. for Aerospace and Petroleum Industries", Indus 14) "High Performance Materials Demand in Aircraft trial Heating, June 1983. Gas Turbine Engines: Market Forecasts, Technol 28) Zecco, J.: Wyman Gordon, Private Communica ogy Assessments and New Business Opportunities, tion, December 1989. 1988-1998", Gorham Advanced Materials Institute. 29) Kulkarni, K.M.: "Isothermal Forging-From Re 14) "High Performance Materials Demand in Aircraft search to a Promising New Manufacturing Tech Gas Turbine Engines: Market Forecasts, Technol nology", Production to Near Net Shape Source ogy Assessments and New Business Opportunities, Book, American Society of Metals, 1983, pp. 157. 1988 -1998", Gorham Advanced Materials Institute.
118 KONA No.8 (1990)