comparação forjado x chapa

8/13/2019 comparao forjado x chapa

1/54

range of components for which plate and forgings com-pete, they represent trends which should be consideredin the material and product form selection process. Thatis, the potential for acquisition cost saving advantage ofthxk plate should be weighed agamst the higher safetymargin and /or weight a nd maintenance cost savingopportunities provided by forgings., IR 2Many s tructural airframe components, such as bulk-heads, spars, landing gear support, etc., manufacturedfrom thick 7050-774Xx forged and plate products arcinternal to the arcraft. These components are subject toless traffic damage than external parts, so that in-servicedamage nucleation is often controlled by intrinsic mate-rial or product features rather than extrinsic events (e.g.,traffic nicks and dings) as is the case of external stmc-ture. Since these thick produc t components are internal,it is difficult to inspect for fatigue damage (structuraltear down is the only reliable method) and repairs can becostly. Because cracks are less tolerable in such internalcomponents, it follows that almost any improvement inthe host (raw) material p roducts durability and damagetolerance capabilities would be embraced by the user,provided sad improvement is shown to be afford-able.24. Considerable research has been devoted to defining therole of rmcrostructure on fatigue crack initiation andgrowth processes in high strength aluminumalloy~.~~~~~~Much of this work has focused on evalua-tion of thick 7050 alloy product forms commonly speci-fied for heavy sectioned internal aircraft components.Extensive fatigue testing has been completed usingsmooth, open-hole and fracture m echanics specimen testmethods to obtain inslght mto the linkages betweenprocessing, microstructure, and performance for the pur-pose of guiding alloy development, raw material productprocessmg development and component material andproduct form selections.Resulting fatigue da ta have been organized for inputinto fracture mechanics based models developed toquantify the benefit po tential of durability-improvedmaterials and to discern the respective roles of metalpedigree, fabricating history, and configuration detail onend-componen t fatigue performance. 4oHigh resolutionfractography and other evaluation techniques performedon failed fatigue specimens have yielded ex tensive sta-tistical data on fatigue life-limiting mlcrofeature popula-tions for several commercial plate and forged prod uctform variants. This work identified a hierarchy offatigue crack mitmting material mlcrofeatures rankedbelow in descending order of severity for thick 7050.174Xx product applications:24(I) Mlcroporosity.(2) Constituent particles.(3) Persistent slip bound aries or other grain structurefeatures.Forged products may be under scrutiny with res pect tothe aerospace industrys drive to reduce manufacturingcosts. However, an understanding of the forgedproducts process-microstructure-performance attributessuggests that forgings in some applications provide theopportunity for performance and maintenance cost sav-ing enhancements. The excellent durability and damage

I. BackgroundPrimary airframe components are designed to static anddynamic loadmg, deformation, and other functional cri-teria. Operating service loading criteria for design andverification of durability and damage tolerance, bothcritical elements of airworthiness, are equally important.However, origination and growth of fatigue cracks haveperpetually challenged airframe engmeers. ) Yet, servicelives of commercial and military arplanes are beingextended due to the increasing costs of fleet replace-ment. As a result, the aerospace industry has increasedits focus on reducing both acquisition costs and totalcost of aircraft ownersh ip. Exploitatmn of aluminumalloys with superior fatigue and durability performancecapabdities provides one means to address both theownership cost and life extension objectives. To obtainthe full benefit potential of these actions, an appropriateaccounting of the aluminum interaction with airframedesign criteria is a necessary prerequisite. This s tep isalso important to planning and execution of cost-effec-tive prevention and control strategies to mitigate poten-tial safety risks associated with airframe aging (e.g., thewidespread damage threat).. xz7From the airworthiness perspective, durability and dam-age tolerance methodologies have three criticalelements:

l Residual strength, a measure of the maximumdamage permissible without impairment to safetyor function.l Crack growth, the interval of damage from non-detection to allowable size determmed by staticstrength safety margins for damaged structure.l Damage detection, the sequence of and frequencyof inspections.

Critical components must be evaluated for these charac-teristics, the outcome of which can affect airframeweight, cost and maintenance. Translation of raw mate-rial and product form characteristics to finished part per-formance according to these metrics is a key element ofthe design and material selection process. yGrowing size of airframes and pressures to reduceairframe build costs have resulted in a trend towardreplacement of built-up assemblies with thick, mono-lithic compon ents. Hand and closed-die forgings andthick plate product forms in high strength aluminumalloys (e.g., 7050 and others) compete for these applica-tions. Some airframe manufacturers are convertingparts traditionally made from forgings to thick platehogouts. Historical technical reas ons for the originalpreference for forged products in critical componentsare being compromised in product se lection delihera-tions to realize a possible manufacturing cost reductionthrough avoidance of machining distortion problemssometimes encountered with forged parts.Structural durabihty and damage tolerance assessmentsbased upon lab-scale test results and limited full compo-nent tests typically project sup erior fatigue perfomxmcefor components manufactured as forged parts versusthose machined from thick plate., I7 Although theseevaluations are limited in scope relative to the broad


2/54

tolerance capabilities of high strength, 70WT74XX part fatigue performance data. Forged products durabil-hand and die-forging emerge directly from the products ity performance is shown to be typically superior tomicrostructure and fabrication (working) history.6, 7 other candidate products. Thus, forgings are expected toThis observation is aligned with classical fatigue be especially exploitable when fatigue and durability areimprovement scenarios, Figure I, the hmlting deslgn criteria. These characteristics areIn this report, forging durability and damage tolerance necessary For the safety, reliability and affordabilityperformance attributes are described phenomen- assurancesdemanded of long-lived, net-shape structuralologically and analytically, using models and coupon vs. airframe parts.


3/54

II. Historical Advantages a nd Performance Capabil-ities of The Forging Process and Hand andClosed-die Forging ProductsForgings have historically been the preferred highstrength, heavy-section, e.g. 2 4 in. thick, aluminumproduct for manufacture of aircraft components becauseof advantagesprovided by the forging deformation pro-cess. Well designed and executed forging andthermomechanical processes (TMP) for alloys such as7050 impart htgh levels of multi-directional mechanicalwork (strain) and microporosity-healing compressivestress throughout the raw forging shape and the finishedcomponent cross-section. By so doing, the microstruc-ture in all grain orientations of the forged product ishomogeneous and thoroughly wrought. The multi-axisforging processes employed develop superior micro-structures and properties in the short transverse (ST)grain orientation and in heavy section thicknesses.In contrast to other conventional mill working processes(e.g., rolling, extrusion, etc.), the forging process, ingeneral, is more robust in its ability to manipulatemicrostructural features controlling key mechanicalproperties. The higher levels of mechanical work typicalof forgings be tter accommodate healing and/or break-upof deleterious remnants of the ortginal cast structure.This deformation reduces both the size and frequency ofweakening microfeatures important to ductility andfatigue strength.Additionally, closed-die forgings provide contour-following, or controlled, grain flows, thereby enablingoptimal tailoring of directional mechanical propertieswith principal load paths in the end-part. Controlledgrain flow with respect to the finished componentreduces end-grain and general corrosion and stresscorrosion cracking concerns. Forged parts with contour-following grain flows will also exhibit superior resis-tance to growth of surface nucleated fatigue cracks, suchas those initiated from corrosion or mechanical damage.Moreover, it is well recognized that forgings offer asuperior combination of damage tolerance and short

transverse ductility capabilities in very heavy sections,e.g. 2 6 in. thick, without excessive strength propertycompromise. Finally, to properly compare forging vs.machined hogout property capabilittes, the comparisonsshould be made on the basis of pre-machined productdimensions. That is. the properties of a reduced forgedcross-section and design are more appropriately com-pared to those of the thicker host plate required to pro-vide the same machined component.The historical property advantages of forged products inatrframe structure are being challenged by other thickaluminum alloy products. However, analytical andmodeling technologies developed from current researchpermit quantification of the durability and damagetolerance performance advantages of forgings. Withthese data on durability and damage tolerance, quantiti-cation of weight savings, life enhancement and totalcost savings from forgings can be accomplished.The following subjects are considered in the balance ofthis report to define the Property and PerformanceAdvantages of Forgings, with Focus on Durability andDamage Tolerance A ttributes:

l Microstructural Basis For Fatigue Performance of7050.T74XX and Other High Strength AluminumAlloys (Section III).. Microstructurally Based Life Prediction Tools(Section IV).l Forged Product Process-Performance Relatton-ships (Section V).. Thick Component Life Prediction; Forgings vs.Machined Plate (Section VI).. Full-Scale Fatigue Evaluation of Prototypical Parts(Section VII).l Other PropettylPerformance Attributes for Aero-space Forgings (Section VIII).


4/54

4

III. Microstructural Basis for Fatigue Performanceof 7050-T74XX and Other High Strength Alumi-urn AlloysEnd-component performance is related to the propertiesof the components starting raw material, which in turnis related to the materials microstructure. Microstruc-ture is defined by the raw material product forms man-ufacturing process (i.e., forging, rolling, machining,etc.). Hence, if the effect of process on microstructure isknown, then material properties and ultimately compo-nent performance are predictable from the governingmicrostructure-property relationships. The availabilityof performance data to compare forgings with otherproduct forms on an equal component basis is limited.However, evaluation of the relative merits of forgingscan be supplemented by first principles logic and under-standing of process-microstructure-performance rela-tionships derived from the extensive work done on otherproduct fomx.2Y-4Fundamental knowledge applicable to assessment of7050-T74XX forgings and other thick product durabilityand damage tolerance attributes was acquired from theseact,v,t,es:

l Quantification of metal pedigree effects on fatigueperfOllllanCe.

l Statistical characterization of life-governingmicrofeatures and their hierarchy of importance forvarious products and specimen types.l Confirmation of fracture mechanics applicabilitydown to the scale of life-limiting microfeatures.l Corrective protocols to remove residual stress biasfrom coupon fatigue crack growth rate data.l Analytical models to define:

- Equivalent Initial Flaw Size (EIFS) from lifedata and random plane microstructures.- Structural life and crack growth probabilitiesfrom EIFS distributions.- Effects of microfeature type, size, shape andlocation distributions.

Critical microstructure-fatigue property relationshipsand hierarchy of life-limiting features for superiorfatigue performance are independent of productfonn.4 I. 2.25-4 owever, as demonstrated from parallelinvestigations, optimization of critical microstructuresand microfeatures for superior fatigue performance isdependent upon product form, and particularly a prod-ucts p rocessing and working history.4~2h~28~3~i7~40~Mtis in this latter aspect that forged products demonstratetheir value. For purposes of the balance of this discus-sion, a life-limiting microfeature is defined fromSection I above as the microstructural feature(s), micro-porosity, constituent particles, persistent slip bands orother Stage I grain structure, identified by high resolu-tion fractography as the initiation site for fatigue failurein smooth or notched specimen fatigue tests.Historical fatigue data from thick 7050-T7451 plate,Figure 2, established a relatmnship between smoothspecimen fatigue life and sizes of the life-limiting

microfeature, in this instance micro-porosity.X~ 3x.4,44,% Metal quality effects on fatiguehave also been confirmed in U.S. Air Force tests onspecimens prepared with production machining prac-tices and representative pan loadings, Figure 3.29 Inthese tests, internal metal quality controlled fatigue per-formance even in the presence of surface-related fea-tures from machining.Initial test programs were focused on 7050.T7451 thick(5.7 in.) plate improvement and identified micro-porosity as the predominant fatigue crack-initiatingmicrofeature, Figure 4. The product quality, in thisinstance, was directly related to the size and distributionof centerline microporosity. Microporosity characteris-tics, in turn, directly relate to fatigue performance insmooth and notched specimens.34-37~y These observa-tions led to the definition of the hierarchy of life-limit-ing microfeatures.*-, I7 Life limiting microfeatureevaluations have subsequently been executed on several7050.T74XX plates and extruded thicknesses from I to6 in.2r-,4.d.M. $ on thick (6.25 in.) rectangular handforging;2 and on several heavy section, large closed-dieforgings, with rib heights and/or cross-section thick-nessesof up to 6 to 9 in.. To mathematically analyze and model fatigue behavioras a function of product quality or form, fatigue datawere statistically treated through cumulative probabilityof failure plots. Flgure 5 presents such a plot and repre-sents a decade of 7050 thick plate process evolution toelevate fatigue perfommnce.*28~3The figure also illus-trates that substantial fatigue benefit is derived from theadded mechanical work of rolling plate to a thin gage .While fatigue design allowables are seldom based onsmooth specimen testing alone, the smooth specimentest is an excellent quality indicator tool (seeks out theweakest material link) as well as a valid tool for pointdesign. Smooth specimen fatigue testing of 7050.T74XX, etaI. is required by some material procurementspecifications as a means of controlling microporositylevels in thick aluminum plate for fatigue criticalpXtS.-8Generation of allowable cyclic operating stresses fordesign generally entails notched specimen testing. Intro-duction of engineering notches into the fatigue testreduces the response to material or product internalquality.d, 4h However, as illustrated in Figure 6, theimpact of intrinsic microfeature characteristics on 7050.T7451 thick plate performance is not erased by theseextrinsic open-hole features. Internal quality measurablyinfluences perfomumce of open-hole coupons and com-ponents they emulate.z-z8~4.. 39The example of Figure6 also illustrates that improving open-hole quality, inthis case by hole deburring, dramatically enhanced per-formance of the process-enhancedmaterial (mid 1980sthrough early 1990s commercial practice), while debur-ring produced little or no improvement in the material oflesser pedigree (1970s through early 1980s commer-cial practice).The above examples illustrate the potential forcompetition among life-limiting microfeatures, and howquality-improved material must be accompanied bymanufacturing improvements to achieve the optimal


5/54

component performance benefits. In the case of thelower quality thick plate material (1980 commercialpedigree), the specimenswith deburred holes yielded thesame S-N fatigue response as specimens with theas-machined holes. That is. coarse microporosityremained the life-controlling feature for the 1980 pedi-gree plate material regardless of whether the hole wasdeburred or not. In contrast, the more refined micro-porosity m aterial (from 1990 commercial practice)exhibited fatigue crack initiation from edge-burrs of theas-machined holes. Upon subsequent esting with debur-red holes, the life-limiting microfeature of theenhanced pedigree plate reverted to either fine micro-porosity and/or material constituent particles.Thus, analysis of microfeatures and their hierarchicaleffect on performance is a necessary component ofunderstanding for material (alloy) and/or productform selections and the design and manufacture offatigue resisting structures. Henceforth, the results ofAlcoa open-hole testing presented in this report is fordeburred holes unless otherwise specified.The developing understanding of crack initiatingmicrofeatures in high strength aluminum has fostered anevolution in thick product quality over the past dec-ade.33-y~4s~udditionally, testing techniques and analyt-ical approaches, both of which have served to betterdefine the nature and hierarchy of microfeatures criticalto fatigue performance, were expanded.2r-3,4 or exam-ple, smooth specimen fractographs in Figure 7 revealthat the size of typical failure initiatmg microporesfound in a low porosity 7050.T745l thick plate variantis greatly diminished from that found in products pro-duced with the commercial fabricating practicesemployed from the late 1980s to the early 1990s. Par-allel fractographic analyses of failed open-hole and othernotched fatigue specimens suggest that as material qua l-ity is enhanced in terms of microporosity refinement,other microfeatures (e.g., constituent particles) competeand evolve as the dominant failure initiators.In general, it has been found that fatigue failuresoriginating from constituent particles result in longerlifetimes than failures originating from micropores.However, it can not be unequivocally stated that life-limiting feature changes from microporosity to smallerconstituent particles will guarantee fatigue improve-ment. As illustration of this point, Figure 8 showsfatigue crack initiating features revealed from failureexammations of deburred, open-hole specimen testsfrom two variants of 7050.T7451 thick (5.7 in.) platepossessing differentiated microporosity refinement. Thelife-limiting feature in the 1990 pedigree product (Fig-ure 8, left) is microporosity just below the surface wallof the drilled hole. In contrast, the life-limiting featureof the lower porosity plate variant (Figure 8, right) is ahard constituent particle located at the hole corner. Thelife-limiting particle in the specimen on the right isabout half the size of the hfe-limiting micropore in thespecimen on the left; however, both specimens failed atcomparable lifetimes.Probabilistic fracture mechanics concepts have aided inunderstanding thx ohservation.3 iI The explanation isbased on the mteraction of the stressconcentrating notch

with competing life-limiting microfeature (e.g., micro-porosity vs. constituent particle) distributions in themost highly stressed material volume (i.e., the metalsampling effect).J7 Because of high volume fractions,constituent particles almost always can he found in thevicinity of hole corners. If large enough, fatigue failureis most likely to originate at the particle-hole cornerinterface where stress concentration is 10.15% higherthan in the hole bore. By comparison, microporoslty ISsparse. Even though life-limiting micropores are largerthan constituent particles, encountering a criticalmicropore at a hole corner is a more rare event. Thus,micropore initiated fatigue failure is more likely to beassociated with the surface region in the bore of the holewhere the probability of a large pore encounter isgreater.The broader implication of these observations is thatwhile smooth specimen fatigue response is more likelyto be controlled by microporosity concentration, notchedspecimen test response s more sensitive to distributionaldifferences among competing life-limiting features andtheir interaction within the highest stressedmaterial vol-~rne.~ Moreover, it is entirely concavable that fatigueratings derived from various types of notched speamentests could be affected depending upon whether the vol-ume of highly stressed material associated with thenotch is large or smalL4The level of thick p roduct microporosity may be refined,for example the low porosity 7050 thick plate variant inFigure 8 and/or by higher deformation levels typical ofclosed-die and hand forged products. In these examples,constituent particles, not microporosity. arc more likelyto become the life-limiting material microfeatures, par-ticularly so in configurations where point stress concen-trations exist (e.g., at hole corners and sharp notches).Figure 9 illustrates that the size and distribution of con-stituent particles within the bore of an open hole speci-men are comparable for the two pedigrees of 7050 thickplate in Figure 8. It follows from Figures 8 and 9, thatenhancement in open-hole specimen fatigue strengthmay not he attained by microporosity refinement alone.Process modification, via working history, compositionor both, to reduce both size and frequency of constituentparticles must accompany microporosity refinement toachieve further open-hole cyclic life enhancement.In contrast, many finished airframe parts contain low-stress concentration cutouts, fillets, lands, etc. for whichthe volume of highly stressed surface material can belarge. In this instance, the fatigue benefit potential ofmicroporosity refinement, achieved primarily fromworkmg history, can be appreciably greater, because alarger volume of maternal IS being sampled than thatobserved from the open-hole specimen result, wherefatigue originates at the hole corners.Figure 5 presents a comparison of lives of three pedi-grees of thick (5.7 in.) and one pedigree of thin (I in.)7050.T7451 plate. All pedigrees are compositionallysimilar. The deformation history of the thin plate prod-uct is different from the three thicker product gauges.Figure IO illustrates fractographs of typical microfeatureinitiation sites in the more highly worked thin (I in.)plate. Note in the thin product, working was sufficient to


6/54

6result in exclusively constituent particles as the life-lim-iting microfeature.Comparing size of the failure originating particles fromFigure 8, right, for the low porosity thick (5.7 in.) platevariant to Figure IO for the I in. thin plate variant, it isclear that the processing (working) history of the thin(I in.) plate provided a more refined life-limiting parti-cle, in this case a particle 50% smaller than that of thethick (5.7 in.) plate. Accordingly, the vastly superiorsmooth specimen fatigue resistance of thin plate shownin Figure 5 is evidence that the higher working receivedby the thin plate is effective as a microporosity suppres-sant. The superior performance of the thin (I in.), morehighly worked plate is a reasonable surrogate for andpredictor of the performance of forged products, particu-larly close-die forgings. Similar working history effectson life-limiting features (constituents) hold for forgedproducts. As discussed below, this is due to the highlevels of compressive stress and mechanical work(strain) imparted by the forging process.The final hierarchical level of fatigue life-limiting mate-rial microfeature is illustrated in Figure 1 As materialand part manufacturing quality is enhanced by process-ing to reduce size and frequency of micropores, particlesand/or manufacturing imperfections (e.g., scratches andtool marks), the default mechanism becomes the moreclassical Stage I slip induced Initiation behavior. Thismechanism has been the subject of much classicalresearch on the subject of fatigueP*,49and is in partexplained by the development and stabilization of per-sistent slip boundaries (PSB)50 and other phenomenabeyond the scope of this report. Post-test ractography ofliterally hundreds of failed smooth and open-hole cou-pons from various thick commercial aluminum productforms, however, has revealed that it is far more commonfor fatigue cracks to initiate from pore or particle typeirregularities, acting singly or interactively, than it is to

have classical Stage I initiated fatigue. That is, the StageI initiation mode represents the ultimate uppa perform-ance bound for commercial thick aluminum product.In summary, the quantification of m icrostructural influ-ence on laboratory investigations of thick aluminumproduct fatigue performance leads to the generalisedhierarchy of life-limiting microfeatures listed below inorder of decreasing importance.* Available researchwork has confirmed that the hierarchy of m icrofeaturesis independent of product form. 15I 26 l. 13. 7. 11.1. 4 ,-hemicrostructurally based knowledge form a set of firstprinciples that can be exploited for analysis and quanti-fication of the part performance of competitive products,including forgings, and the related effects of m icrostruc-ture, processing and part making histories characteristicof the uses of these competitive products.The essential points to be taken from the discussion ofthis section are that(1) In absence of full compon ent fatigue test results,

the technical concepti of life-limiting microfea-tures and various analytical approaches tofatigue, durability and damage tolerance evalua-tions can be used to understand advantages offorgings. Forg ed products derive their superiorfatigue strength a nd durability characteristicsfrom working and process histories that enablealteration in size and frequency life-limitingmicrofeatures in ways unavailable to otherwrought product forms.(2) Optimal transfer of laboratory data to fatiguedesign is possible when the role of life-limitingmicrofeatures is understood. The accuracy ofprojected outcomes ca n be affected depending

upon the degree of control of said feature(s) inmanufacturing and the quality in the detail of theplanned evaluation.


7/54

7

employed for this calculation is presented in Figure 12and good agreement of the model wtth experimentalobservations has been demonstrated.hFlaw populations from different life-limiting feature dis-tributions, Figure 14, may then be used to assess theeffect of material manufacturing processes,compositionand other process variables on structural component per-formance. In Figure 14, the three flaw size populationsrepresent the evolution of product quality in a selectedthick product, or alternatively, the input EIFS distribu-tions of three competitive product forms. The perform-ance models developed in References 7, 8, 35, and 36enable the frequency of flaw size excedance(s) o be pre-dicted for prototypical structural components.The prediction of flaw size excedances s accomplishedby computing the distributional crack growth accruedfrom simulated in-service loadings and presumed tooriginate from the starting EIFS population. The EIFSpopulation is in turn linked to raw material microstruc-ture. The hypothetical calculation of Figure I4 wasundertaken to illustrate how three representative inputmaterial qualities or pedigrees, based upon identifi-able initial flaw popu lations linked to material micro-structure, might affect the potential for multi-site fatiguedamage (MSD) in the lower wing of a mtlitary tighter.Detailed discussion of this model, and parallel modelsutilized in analyses of other service performance crite-ria, is beyond the scope of this document. References 7,8, 23, 30, 31, 35, 36, 50 and 56 are recommended toprovide more detail on mode ling techniques correlatingflaw population predictions with component or systemperformance through durability and damage tolerancemethodologies.The basic conclusion to be drawn from this discussion isthat reasonable estimates of durability and damagetolerance c apabilities of actual components can belinked with and modeled from life-limiting microfea-ture populations derived from laboratory coupontests. The models p rovide a set of analytical tools totranslate understanding of life-limiting features ofmaterial microstructure, known to emanate from thecompositional and processing history of any givenaluminum alloy product, into useful design guidancefor airframe manufacturers. Such models can be andhave been successfully used to assist in componentdesigns, evaluations of aluminum raw material productoptions and raw material product quality enhancementefforts. 38. .41

IV. Microstructurally Based Life Prediction T oolsSubstantial recent progress has been made in developingfracture mechanics based tools to relate microstructureand fatigue performance.~ %3. b. d.


8/54

8V. Forged Product Process-Performance Relation-shipsIn Section III, a microstructural basis and hierarchy oflife-limiting microfcaturcs controlling smooth and open-hole (notched) fatigue performance in 7050.T74XX (orother aluminum alloys) was established. Descriptions ofequivalent initial flaw size (EIFS) populations linked tolife-limiting microstruc tural features are then utilized asinput mto fracture mechanics models to predict crackorigination and growth in end-components (Section IV).Calculated results from these technologies have beenshown to agree favorably with results of actual structuralelement tests. Thus, an expanded number o f material,product-form and manufacturing options can be evalu-ated analytically, at much earlier stages of design, with-out having to resort solely to prohibitively large andexpensive test programs. In this section, the capabilitiesof candidate aluminum raw material products for themanufacture of components are quantified to facilitateselection on the basis of performance, costs or a combi-nation of both.In selecting among material and product form optionsfor m anufacture of fatigue critical components, from theabove analyses, it is entirely conceivable that workinghistory differences among competing products will havea bearing on acceptable end-component performance.For example, consider a typical large, web-rih compo-nent, whose finished cross-section is shown as ghostoutlines in Figure 15. This part may be produced eitheras a hogout from the thick plate or as a finishedmachined part from the closed-die forging depicted inthe figure.Figure I5 delineates the significant differences in defor-mation imparted to the two candidate host raw materialproduct form options: (1) a thick plate 7 in. x 37.25 in.(cross-sectional area 260 sq.in.) and (2) a large closed-die forging with webs thickness at 0.75 tn. and ribheights of 4.75 to 6.75 in. (cross-sectional area 70sq.in.). Plate 7.0 in. thick is necessary o assure part cov-erage by allowing for all adverse tolerances. From thecontoured shape of the die forging, maximum rib heightis reduced to 6.75 in. for machined part coverage due toreduced adverse tolerance build-up. While comparisonplate thickness versus die forging web thickness is not awholly accurate measure of the working differencesbetween the two products because of the multiple axisworking utilized in forgings, it is a reasonable generalIndicator. On the basis of plate thickness vs. die forgingweb thickness, the deformation (e.g. reduction or levelof strain) of metal into the webs of the die forging is atleast nine (9) times greater than thick plate. However,comparison of plate thickness vs. die forging web thick-ness is not a reasonable gauge of differences in micro-porosity-healing compressive stress introduced. Theforging process has been shown5.. 17.z7o impart veryhigh levels of compressive stresses to the work piecethat are not imparted by the rolling process.From the preceding discussion (Sections III and IV), it isquite reasonable to expect that fatigue testing wou ldreveal differences in raw material intrinsic life-limitingmicrofeature populations traceable to the differentiated

metal working experience of the two host products inFigure 15. Specifically, webs and rib fatigue perform-ance in the end-component is expected to be controlledby constituents or PSBs and grain structure in the moreheavily workcd forging, while constituents or micro-porosity are the life-limiting features anticipated for thethick plate hogout. Such expectations are confirmed byactual observations, as discussed below.Plate fattgue performance is known to be affected byproduct thickness, as illustrated in Figure 16 for smoothspecimen tests at maximum cyclic stress of 40 ksi. Forthick plate, deformation during rolling may be insuffi-cient to hea l centerline microporosity present in thestarting ingot. However, plate produced with an optimalmicrostructure, low porosity 6 in. plate in Figure 16, hasclearly superior fatigue performance. Increasing theamount of mechanical work in rolling to thinner plategages decreases both the amount and size of micro-porosity, with a commensurate smooth specimen fatigueimprovement. The plate gage dependency persists togages ess than 3 in., Figure 17. In the latter cases, how-ever, it is necessary to increase the test stress level to45 ksi (about 65% of yield strength) to differentiatesmooth fatigue performance among the different platethicknesses.Process and working histories of hand forgings, closed-die forgings and plate are uniquely different. Becausethese histories impact pore and particle populations andgrain morphology in different ways, it follows that rela-tive fatigue performance ratings among plate, hand forg-ings and closed-die forged p roducts will be affectedwhen these features are the life-controlling factors, Fig-ure 18. Note that the most recent quality evolution ofoptimally produced, low porosity thick plate comparesfavorably in performance with thick hand forgings,whose pedigree dates to the early 1990s. I?te early1990s pedigree closed-die forgings remain superior tothe optimal plate product. Capture of the same mprove-ments employed in the most recently produced (1995)optimal, low porosity plate product can be anticipated toimprove performance of hand and die forgings as well.Heavy-section forged products exhibit the same types oflife-limiting microfeatures as thick plate, Figure 19.However, the forging process appears to modify(reduce) the size and frequency of micropores, reducethe size of constituents and drive the life-limitingmicrofeature to PSBs. For example, compare the veryfine mtcroporosity observed in Figure 19, pore lengthsof 0.0006 to 0.0009 in., for specimens taken from aheavy section 7050 die-forging with maximum thick-ness of 6.3 in. vs. the life-limiting microporosity sizesobserved in Figure 4 (0.002 to 0.004 in.) and Figure 7(0.005 to 0.0015 in.) that illustrate evolving quality of7050 thick plate.Fractographic examination of numerous failed smoothand notched fatigue specimens extracted from varioustype forgings (thick and thin)lb, reveals that constituentparticles and Stage I PSBs (comparable to Figures IOand I I, respectively) compete as the most frequentlyencountered type of material-intrinsic crack initiatingmicrofeature in forgings. In contrast to thick plate where


9/54

microporosity initiated fatigue cracking is a more corn-moo occurrence, the more effective microporosity heal-ing achieved by the forging deformation history drivesthe fatigue process to seek a lower ranking member ofthe bfe-lmutmg maofeature cham, Table I.These microstructural observations for forgings confirmthat the working history of forged products is the over-powering feature of this raw material form in terms ofits fatigue performance advantage over plate, where therolling process does not introduce high levels of com-pressive stress. The fatigue lifetimes of specimens fromheavily worked die forgmgs even exceed those of speci-mens removed from thin plate, Figure 20. In thesetests, all specimens excised from the web section of adie forging with a rib height of 6 in. (comparable platewould be at least 6 in. thick) were runout.$at stressesofat least 60% of the yield strength of 7050-T74XX, dem-onstrating the superior fatigue strength of forgmgs.A forger has several options available in design of theforging process for any given forged part, but particu-larly for large, heavy section closed-die forgings. Theseoptions include (I) selection of input material, e.g.as-cast ngot vs. wrought stock forms of rolled, drawn orextruded bar or plate; (2) selection of preform or open

die forgmg processes and preform geometry; and (3)selectmn of single or multiple closed-die working pro-cess histories. From the analyses and evaluationspresented above, starting stock, pre-working and finaldeformation process(es) selections may have markedinfluence on the final forged (and machined) partsfatigue performance. This observation is illustrated inFigure 21, where Process Route B produces clearlysuperior smooth specimen fatigue performance overProcess Route A.h. The improvement of performancethrough Process B IS attributed to a multiple-axis work-ing hlstory, with higher levels of total forging reductionthan Process A parts, which have a single axis deforma-tion history.Finally, the performance advantage of forgings, m par-ticular die forgings, is retained even in the presence ofnotches, as dlustrated in Figure 22. Plotted are open-hole fatigue results from the same plate and forged prod-ucts (Process B) presented in Figures 20 and 21.In summary, the forging process itself affords ameans to further differentiate the fatigue pe rform-ance advantage of forged parts from that of compa-rable thickness plate hogouts.

Table 1: Hierarchv of fatigue crack initiating microfeatures commonlyobserved in coupon specimen evaluations of commercial 7050-T74XX thick products.Predominant life-limiting microstructural feature

g&(1)(2)(3)(4)

Smooth(round bars)Coarse microporosity

MicroporosityConstituent particles

Grain structure

Open-hole(as-machined)

Coarse microporosityMicroporosity &hole quality

Hole quality

Open-holeJdeburredj

Coarse microporosityMicroporosity

Constituent particlesConstituent particles & grainStruUCtUrc


10/54

VI. Thick Component Life Prediction; Forgings vs.Machined PlateFull component test appraisals of the performance ofcompetitive high strength aluminum product forms arecostly. Thus, few have been undertaken and reported,other than those discussed n Section VII below for pre-cision forgings. Because it is necessary or the airframemanufacturer to make quantifiably sound componentdesign and material product selection udgm ents, predic-tive models facilitate an expanded range of options, inshorter times, and at lower costs.The majority of high strength aluminum alloy compo-nent durability and damage tolerance predictive model-ing has been based upon 7050 alloy plate fatigue crackgrowth rate (FCGR) performance.25-.34-6n this section,justification for using plate fracture mechanics data andmodels for forgings, with appropriate treatment of possi-ble residual stress bias in forging FCGR data generation,is p resented. It is shown that forged product capabilitycan be reasonably estimated from available plate FCGRproperty data as a function of grain orientation in orderto demonstrate the benefit potential of forgings con-tour-following grain flow and refined life-limitingmicrofeature population. These model results are thenreviewed in terms of the potential life extension, weightand cost reduction advantages of forgings.Life prediction models begin with fracture mechanicsanalyses based upon the material FCGR relationshipexpected for the end-component. Insertion of forgingdata into plate models and utilization of plate data toanalyse forging product performance requires justitica-tion. Previous published work has addressed he fracturemechanics performance issues of this cower-sion.12. I. 15. 7Plate FCGR properties for 7050.T745l have beenshown to be independent of product thickness, Figure23, and the FCGR properties of 7050-T74XX in forg-ings and plate are comparable, compare Figure 23 toFigure 24 . However, the industrys FCGR database for7050 forgings is tainted by test method induced biases,e.g., Figure 25, that produce spurious FCGR measure-ments from coupons isolated from partially stressrelieved parts., lr. 16. 2A method has been developed5to remove the residual stress bias from FCGR data gen-erated by standard fracture mechanics test methods, Fig-ure 26.The correction practice employs a factor, derived fromcrack closure-type measurement techniques, to removeinternal residual stresses rom computation of the speci-men cyclic crack drive (AK) solution. Without this COT-rection, the residual stress bias introduced in the couponhas been shown in some cases (dependent on specimenand part geometry) to produce an effect on crack drivingforce that is greatly exaggerated beyond that observedfor naturally originating cracks in actual parts.2A comparison of Figures 26 and 27 shows that when thesame crack closure-based correction practice is appliedto coupon generated data from stretched 7050 plate andnon-stress relieved 7050 forgings, the FCGR relation-ships are similar. Hence, the intrinsic FCGR perforn-ante of forgings can be reasonably estimated from plate

data of comparable grain flow.~ I6 Thus, when forgingFCGR data are either not available or of suspectquality,reasonable first estimates of forging FCGR performancecan be judged from plate FCGR perfomxwce.An optimally designed web-rib die-forging requiresminimal finish machining so as not to disrupt the con-tour-following grain flow in the final part. Considerthe engineering component of Figure 28, which con tainsattachment holes and can be manufactured either as aplate hogout or a near-net-shape die forging. In the die-forging, the predominant grain orientation in the ribchanges with part contour in comparison to a partmachined from plate, where grain flow cannot betailored. The die-forging grain flow may be controlledto be favorably oriented with the part load path, com-pared to that in a plate component where machiningmay interrupt the predominantly longitudinal plate grainflow. FCGR performance of plate, which will be utilizedas an indicator of FCGR in forgings, varies with grainorientation (Figure 28 inset), even when corrected forcrack closure, Figure 29. If one considers crack LocationA in Figure 28, the plate grain orientation to establishFCGR perfomumce is S-L, whereas in the forging it isL-T, due to conformance of the grain flow to the com-ponent shape.The potentially substantial fatigue benefit of forgings isdemonstrated by the results portrayed in Figure 30. Thefigure presents the predicted open-hole S-N fatigueresponse of a hypothetical part (like that in Figure 28)machined from plate versus the same part designed as anoptimal die forging. The calculation employed a genera-lized probabilistic crack growth model with power lawFCGR relationship (extended to the short crack regime).Cracks were assumed to initiate from the respectiveproduct life-limiting microfeature populations (insetFigure 30), and the model considered the possibility thatcracks could occur anywhere along the bore of the bole.The input FCGR behavior for the model was derivedfrom plate test results. Crack growth was presumed to bein the L-T and S-L orientations for forging and plate,respectively (i.e., analogous to location A in Figure 28).The upper and lower S-N bounds for the two competi-tive products illustrate the maximum possible differencein closed-die forging vs. plate hogout S-N response forthe modeling assumptions used.The hypothetical examples of Figures 31 and 32 furtherillustrate the potentially substantial weight and/or opera-tor cost saving opportunities afforded by the forgingprocess. Figure 31, aga in contrasts the calculated fatigueperformance of a part machined from plate against thatof the same part designed as an optimal d ie forging. Thecalculation employs a deterministic crack growth model,assuming the fame FCGR relationship assumed for thecalculated results of Figure 30. Cracking is presumed togrow from an open hole starting as a corner crack equalto a representative mean life-limiting feature size of therespective raw material product form (0.003 in. for forg-ing and 0.004 in. for plate).Under the above set of assumptions, the forged partaccommodates3 I % higher maximum cyclic stress (fromIO ksi to 13.1 ksi) at equivalent fatigue life for p late, orthe die forging can accommodate about a five times life


11/54

(3) These advantages for forgings can best be real-ized when p lanned for at the outset of the designprocess and entail a concurrent engineeringapproach, teaming the forging supplier and theSW.Premature specification of an overwed forging toaccommodate uncertainties in design and productionmachining scheduling may entail heavier metal removalthan necessary, negatmg the potential advantage offorged products working history and contour followinggrain flow. Fabrication of an opttmized forged part, par-ticularly closed-die forgings, with superior fatigue per-formance can best be accomplished when the finalmachined parts geometty requirements, and therefromthe forged parts final geometry. have been fullyresolved in design. If not fully resolved in design, beforefabrication respective final part and forging designsmust be as close as possible to this point in the compo-nents manufacturing history. Concurrent engineering isthe technology which facilitates this objective.

increase over plate for the same IO ksi maximum cyclicoperating stress. A companion example to Figure 31 ispresented in Figure 32. In this latter example, crackgrowth is presumed to start from a mechanical 0.010 in.flaw located at the corner of an open hole. Under thebest possible scenario, where crack growth is in the L-Tvs. S-L orientation for forging and plate respectively,the die forging accommodates a 27% higher cyclicstress (from IO ksi to 12.7 ksi) at equivalent fatigue lifefor plate, or almost a four tmxs life increase in life forthe die forging over plate for the same IO ksi maximumcyclic operating stress.In summary, as the preceding examples illustrate:(1) The fatigue advantage of forgings can be morethan incremental, provided the positive attributesof forgings are tailored to the final partobjectives.(2) The preceding hypothetical calculations showthat when forgings are tailored to final part

objective, near-net-shape die forgings, in particu-lar, can translate into opportunities for substan-tial safety margin increase, weight reductionand/or inspection interval increase.


12/54

VII. Full-Scale Fatigue Evaluation of PrototypicalPWt.9Ideally, manufacturers would like to have data from fullcomponent tests of critically selected structural elementsto appraise the competitive and airworthiness aspects ofvarious material product form options. With actual partresults, durability and damage tolerance models for thevarious manufactured part options could be validateddirectly, and the data would be useful for extending theselection criteria to other component scenarios. How-ever, side-by-side full component tests of candidate rawmaterial products fabricated into the same finished com-ponents are both costly and rare.One such study was undertaken to determine the relativeresistance to fatigue resulting from aluminum IXXXalloy net-shape precision forging and hogout technolo-gies. Excerpts from this study arc presented in Figures33 to 35. The program tested fatigue performance of fin-ished components instead of specimens cut from theparts themselves, Figure 33. The precision forgings weremanufactured and heat-treated to blueprint specifica-tions. The hogout parts were machined to final dimen-sions from certified heat-treated plate or extruded bar. Anumber of the components were also tested with andwithout surface shot peening. A typical result for one ofthe configurations evaluated is shown in the cumulativefailure probability plot, Figure 34. while Figure 35 sum-marines data from all configurations tested. Both figuresclearly display the superior fatigue resisting capability

of the precision forged parts evaluated, which on aver-age doubled the lifetime of their equivalently testedhogout counterparts.Similar full component fatigue tests of parts machinedfrom heavy section hand and closed-die forgings andparts machined from thick (2 4 in.) plate have beenrecommendedR-22ut have not yet been conducted. Inthe absenceof such full scale test data from thick prod-ucts, critical appraisal and acceptance of the hand andclosed-die forging products fundamental durability per-formance advantages rest upon the following train o flogic:

l The forging product has inherently smaller andlower ranking life-limiting microfeatures due tothe products fabrication and working history.l Hand and closed-die forging products superiormicrostructure translates directly, through durabil-ity and damage tolerance models, to weight sav-ings from higher allowable operating stresses for

equivalent lives and/or life cycle cost savings fromlonger inspection and maintenance intervals.l Airworthiness of components manufactured fromforgings are superior in terms of higher residualstrength at comparable lives, longer intervalsbetween inspections and improved manageabihtyof multiple-site fatigue damage (MSD).


13/54

13

grain than components made from other productswhich can:- Prevent or delay stress corrosion cracking and

other forms of intergranular attack.- Impede corrosion damage penetration and

reduce size, severity and spread of potentiallyfatigue life-limiting microfeature[sl, such asintergranular fissures and pits.- Delay or retard fatigue crack growth originating

from corrosion-cause d microfeatures.C. Other Forging Design and Evaluation Issues:Some development is needed to capture the full potentialof forgings in this era of focus on affordability, lifeextension, and supportability:

l Forgings become more attractive when pointdesign approaches are considered instead ofallowahles and empirical factor-basedapproaches.

l Generic data and efficient analytical procedures areneeded to reduce the expense of part design andcertification on a case-by-case basis.

l The role of microstructu re and its relationship todesign and final part certification must be undcr-stood and integrated into the design and certilica-tion process.

l Good test practices must be followed to minimizethe scatter effect.

. Enhanced controls are needed in forging manufac-turc to assure consistent, high quality, and repro-ducnble end products.

l Component techmcal demonstrations must bedeveloped to serve as validation of the buildingblocks and models.

l Optimum forging utilization requires customer andsupplier relationships that integrate forging attrib-utes into the design and manufacturing cycle,through techniques such as concurrent engineering.

D. Manufacturing Issues Being Addressed in ForgedProducts: The aerospace industrys emphasis on acqui-sition cost, lead-time and flow-time reductions havebrought into focus several issues in the manufacturingstream of forgings that affect the hand and closed-dieforged products user friendliness. The issues beingaddressed are:

l Machining performance enhancement throughcapture of advanced immersion quenchtechnologyr4. Ti; quench path, stress rel~eval andmachining modeling and process optimiration; andoptimal metal removal planning.

. Cost, lead-time and flow-time reductions on bothtooling and forged products through elec tronic datainterchange; computer aided engineering. artificialintelligence and expert system s; fimte elementmethod process modeling; and Concurrent Engi-neering.

l Product consisten cy through statistical p rocesscontrol and statistical quabty control.. Product and process change mana gement.

VIII. Other Property/Performance Attributes forAerospace ForgingsA. Static Strength Properties: Static strength remainsan important design characteris tic and measures the loadcarrying capability of undamaged struc tures. Gene rally,static properties are the subject of allowables hand-books, either industry-wide or customer-specific. Anumber of points regarding the static strength propertiesof forgings should be noted:

. Forgings have superior short transverse ductility(Figure 36) at strength levels comparable to otherthick products of equivalent gages.

l To properly compare properties of shaped die-forg-ings against those of machined plate hogouts, theproperties of the latter sh ould be based on productshaving adequate gage thickness to machine theend-part.

. Handbook design mechanical properties tend to heunintentionally biased against forgings because ofa number of reasons asso ciated with data collec-tion, data reduction, statistical manipulation andsubtle differences in the presentation format. Forexample:- Handbook propeny values for forgings are typi-

cally obtained from pooled data that consoli-dates widely variant shapes and fabricationhistories. The resulting statistical scatter undulypenalizes design property values attached to cer-tain forgings. A more level playing field forforgings is created when specification andallowable property determin ations are based onproduct(s) with well defined fabrication historyand fewer shapes.

- Die-forgmg transverse (not-parallel) handbookvalues integrate the an&g of LT, ST and 45grain orientations for other products (e.g., plateand extrusion) and are therefore very conserva-tive due to scatter inc rease in forging data. Forimproved comparison, forging allowables wouldbe better based on specification s and orienta-tion(s) representative of actual grain flow(s) inthe end-part.

- In Mil-Handb ook-& bearing strength propertiesof forgings may be penalized by up to 15% vs.other products (plate) due to differences in spec-imen orientation. Rectification of this difference,includmg conversion factors, is being sought.

B. Corrosion Performance: Aging a ircraft issues andlonger serwce lives have increased the emphasis onenhanced lon g term corrosion performance of structuralcomponents and the materials from which they are man-ufactured. The following points concern corrosion ofthick forged products:

. Hundreds of tests .on 7050.T74XX (and other7XxX-T7X type alloy) forgings performed overmany years have shown no difference in corrosionperformance for forgings vs. other products inequivalent grain orientations.~

l Closed-die forging proce sses and properlydesigned closed-die forgmg shapes have part con-forming grain flow that results in less exposed end


14/54

IX. ConclusionsFrom data and results presen ted on: (I) microstructu ralbasis for fatigue performance of 7050.T74XX and otherhigh strength aluminum alloy forgings; (2) microstm c-turally based life prediction tools; (3) forged produ ct-process performance relationships; (4) thick componentlife prediction of forgings vs. machined plate; and (5)full-scale fatigue evaluation of prototypical parts, thefollowing conclusions can be reached:A. Aluminum hand and die-forgings for aerospaceapplications can be differentiated as a high qual-ity product with performance benefits that cancreate user preference.B. The fatigue and durability performance advan-tages of 7050-T74XX and other high strengthaluminum alloy forged products re directly cur-relatable to the reduced size and population (orfrequency) of characteristic life-limitingmicrofeatures.C. The reduced size and population of life-limitingmicrofeatures in forgings re fundamentallylinked with the working histories employed in

forging fabrication. Forged product consistencyin durability and fatigue performance, and possi-bly other properties, is linked with selection andexecution of optimum working histories andstarting stock.D. Microstructurally based durability and damage

tolerance models enable the merits of forgings tobe conveyed to final airframe components interms of: weight sa ving potential from higheroperating stress allowables at equivalent lifetime;maintenance cust savings from longer inspectionand maintenance intervals at comparable operat-ing stress levels; and/or rdlowable increase in ser-vice lifetime beyond existing design envelopes.E. Parts manufactured as optimally designed forg-ings offer superior airworthiness capabilities interms of higher structural residual strength capa-bility at comparable lifetimes, longer intervalsbetween inspections, and improved preven-tion/control management of widespread d amageand its consequences.


15/54

REFERENCESI. Airworthiness Standards: Transport Category Air-planes 14 CFR part 25 of the Federal AviationRegulations (FAR), Department of Transportation,Federal Aviation Administration.2. Department of the Air Force, Aircraft StructuralIntegrity Program Requirements, Military StandardMIL-STD-1530A, 1975.3. U. G . Goranson and J. F. McGuire, Airworthi-ness, Aerospace Engineering, March 1994, pp. 28.34.4. U. G. Goranson, Elements of Structural IntegrityAssurance, Fatigue of Aircraft Materials, Proceed-ings of the Specialists Conference dedicated to the65th Birthday o f J. Schijve, Delft University Press,the Netherlands, 1992 pp. 275.320.5. U. G. Goranson, Damage Tolerance Fact andFiction, 14th Plantema Memorial Lecture, 17 th

Symposium of the International Committee on Aer-onautical Fatigue, Stockholm, Sweden, June 9,1993.6. A . P. Berens, D. S . Morcock, K. E. Brown, J. C.Davis, R. L. Johnson, C. E. Larson and D. J. White,Handbook of Force Management Methods,AFWAL-TR-81-3079, Flight Dynamics Laboratory,Air Force Wright Aeronautical Laboratories, AirForce Systems Command, W right-Patterson AirForce Base, OH, April I98 I.7. J. L. Rudd, J. N. Yang, S. D. Manning, B. G. W.Yee, Probabilistic Fracture Mechanics AnalysisMethods for Structural Durability, in Behavior of

Short Cracks in Airframe Components, AGARDConference Proceedings No. 328, NATO AdvisoryGroup on Aerospace Research and Development,April 1983, pp. IO-I 10.23.8. S. D. Manning and J. N. Yang, USAF DurabilityDesign Handbook: Guidelines for the Analysis andDesign of Durable Aircraft Structures, AFWAL-TR-83.3027, Flight Dynamics Laboratory, AirForce Wright Aeronautical Laboratories, Air ForceSystems Command, Wright-Patterson Air ForceBase, OH, August 1988.9. T. Swift, Widespread Fatigue Damage Monitoring- Issues and Concerns, 5th International Confer-ence on Structural Airworthiness of New and AgingAircraft, Hamburg, Germany, June 16-18, 1993.

IO. W. Schutz, Design of Heavy Sections, FractureMechanics Design Methodology, AGARD LectureSeries No. 97, AGARD-LS-97, Advisory Group forAerospace Research and Development, NorthAtlantic Treaty Organization, 1979, pp. 7-l 7-l I.I I. Ultra-Thick Plate Used in 767 AWACS Supports,Aviation Week and Space Technology, March 20,1995, pg. 34.12. R.J. Bucci, Evaluation of Aluminum Plate andForging Fatigue and Fatigue Crack Growth Behav-iors, WESTEC 1992, Society of Manufacturing

Engineers, Los Angeles, California, 1992 March.

13. Wetght and Cost Benefits of Precision Forgings,Precision Forging Assoaatlon of Southern Califor-nia, November 1992.14. R.W. Bush. R.J. Bucci. P.E. Maenusen and G.W.Kuhlman, Fracture Mechanics characteristics ofHugh Strength Aluminum Alloy Forainas, AlcoaForged Products, April 26, 1993.15. R. W. Bush, R. J. Bucci, P. E. Magnusen, and G. W.Kuhlman, Fatigue Crack Growth Rate Measure-ments in Aluminum Alloy Forgings: Effects ofResidual Stress and Grain Flow, Fracture Mechan-ics: Twenty Third Symposium, ASTM STP 1189,American Society for Testing and Materials, Phda-delphia, 1993, pp. 568.589.16. R.W. Bush, Effects of Starting Stock on FatigueQuality of 7050.T74 Closed-Die Forgings As Com-pared to 7050.T745X Hand Forgings, Plate andHFL Plate, Alcoa Laboratories, Report 57-93-20,1993 July 9.17. R. J. Bucci, R. W. Bush, A. J. Hmkle, H. J. Konish,M. Kulak, R. H. Wygonik, and G. W. Kuhlman.High Strength Aluminum Forged Product Dumbil-itv and Damaze Tolerance Caoabilities for Weightand Total Life Cycle Cost Sa&g, Proceedings-ofthe 1995 USAF Structural Integrity Program Con-ference Volume II, WL-TR-96-4094, Wright Labo-c;rrPAFB, Ohio, August 1996. pp. 681-702.18. Private Communications, Northrop GrummanCorporation, 1992 March.19. Private Communications, Martin Lockheed Cor-poration, 1993 June.20. Private Communications, Boeing Company, 1994October.21. Prwate Communications, McDonnell-DouglasWest, 1994 November.22. Private Communications, McDonnell-Douglas East,1994 December.23. J. C. Newman, Jr., E. P. Phillips, M. H. Swain andR. H. Everett, Jr., Fatigue Mechamcs: An Assess-ment of a Umiied Approach to Life Prediction,Advances in Fatigue Lifetime Predictive Tech--, ASTM STP 1122, American Society forTesting and Materials, 1992, pp. 5-27.24. A. J. Hinkle, P. E. Magnusen, R. J. Bucci and J. L.Rudd, Fatigue Durability Improvement in Thick-Section Metallic Airframe Parts, White Paper,Appendix, Advanced Metallic Structures Technol-ogy Development Plan, Aerospace Industries Asso-ciation, 1993 March 31.25. R. J. Buca, J. R. Brockenbrough, A. 1. Hinkle,P. E. Magnusen, and S. M. Miyasato, The Role ofMicrostructure on Fatigue Performance of MetallicAlloys, Office of Naval Research,Structural Relia-bility Progress Report Meeting, Reston, VA,December 13, 1993.26. P. E. Magnusen, R. J. Bucci, A. J. Hinkle, J. R.

Brockenbrough, S. M. Miyasato and H. J. Konish,


16/54

The Role of Microstructure on Fatigue Performante of Aluminum Aerospace Alloys, AnnualContract Briefing: ONR Contract No. NOOO14-91.C-0128, Office of Naval Research, Arlington VA,January 20, 1995.27. P. E. Magnusen, R. J. Bucci, A. 1. Hinkle, J. R.Brockenbrough, H. J. Konish, and S. M. Miyasato,The Role of Microstructure on the Fatigue Durabil-ity of Aluminum Aircraft Alloys, Final Report,Office of Naval Research Contract NOOOl4.91-C.0128, November 1995.28. P. E. Magnusen, R. J. Bucci, A. J. Hinkle, J. R.Brockenbrough, and H.J. Konish, Analysis andPrediction of Microstructural Effects on Long-TermFatigue Performance of an Aluminum AerospaceAlloy, First International Conference on FatigueDamage in Structural Materials, Hyannis, MA, Sep-tember 22-21, 1996 (accepted, International Jour-nal of Fatigue).29. A. F. Grandt, T. D. Scheumann, R. E. Todd andA. J. Hinkle, Modelling the Influence of InitialMaterial Inhomogeneities on the Fatigue Life ofNotched Components, Fatigue and Fracture ofEngineering Materials and Structures, Vol. 16,No. 2, 1993, pp. 199-213.30. A. J. Hinkle and A. F. Grandt, Predicting the Influ-ence of Initial Material Quality on Fatigue Life,ICAF93, International Committee on AeronauticalFatigue, Stockholm, Sweden, June 7-11, 1993.31. J. R. Brockenbrough, A. J. Hinkle, P. E. Magnusenand R. J. Bucci, Microstructurally Based Model ofFatigue Initiation and Growth, FAA/NASA Inter-

national Symposium on Advanced Structural Integ-rity Methods for Airframe Durability and DamageTolerance, NASA Conference Publication 3274,Part I, September 1994, pp. 71-84.32. T. H. Sanders, Jr. and J. T. St&y, Review ofFatigue and Fracture Research on Hirh StrengthAluminum Alloys, Fatigue and Mic~ostruct&,American Society for Metals, Metals Park, OH,1979,p. 467.33. C. R. Owen, R. J. Bucci and R. J. Kegarise, AnAluminum Oualitv Breakthrounh for Aircraft Stmc-tural Reliab%ty,Journal of &xaft, Vol. 26, No.

2, Feb. 1989, pp. 178-184.34. A. J. Hinkle, P. E. Magnusen, R. L. Rolf, R. L. andR. J. Bucci, Effect of Microporosity on NotchedSpecimen Fatigue Life, Structural Safetv and Reli-m, Vol. 2, Proceedings ICOSSAR89, FifthInternational Conference on Structural Safety andReliability, American Society of Civil Engineers,New York, NY, 1989, pp. 1467.1474.35. P. E. Magnusen, R. J. Bucci, A . J. Hinkle, M. E.A&y and R. L. Rolf, The Influence of MaterialQuality on Airframe Structural Durability,Advances in Fracture Research, Vol. 2, Proc. Sev-enth International Conference on Fracture (ICF 7),

1969 March 20.24, pp. 999.1006.

36. P. E. Magnusen, A. J. Hinkle, W. Kaiser, R. J.Bucci and R. L. Rolf. Durabilitv AssessmentBased on Initial Material Quality, J&al of Test-ing and Evaluation, Vol. 18, No. 6, Nov. 1990,pp. 439-445.37. P. E. Magnusen, A. J. Hinkle, R. L. Ro lf, R. J.Bucci and D. A. Lukasak, Methodology for theAssessment of Material Quality Effects on AirframeFatigue Durability, in Fatigue 90, Vol. II, Proc.Fourth International Conf. on Fatigue and FatigueThresholds, Materials and Com ponents EngineeringPublications Ltd., 1990, pp. 2239.2244.38. A. J. Hinkle and M. R. Emptage, Analysis ofFatigue Life Data Using the Box-Cox Trans-formation, Fatigue and Fracture of EngineeringMaterials and Structures, Vol. 14, No. 5, 1991, pp.591.64lo.39. P. E. Magnusen, R. J. Bucci, A. J. Hinkle, J. G.Burns, J. L. and Rudd, Effect of Microporosity on

Fatigue Durability of Thick 7050 Aluminum Plate,Proceedings of the 1992 USAF Structural IntegrityProgram Conference, WL-TR-93-4080, MaterialsDirectorate, Wright Laboratory, Air Force MaterielCommand, Wright Patterson AFB, OH, September1993, pp. 379.410.40. D. A. Lukasak, B. J. Sova and R. L. Lash, ThickProduct 7050.T74511 Extrusions, AeroMat94,ASM Internationals Advanced Aerospace Matern-als/ProcessesConference, Anaheim, CA, June 6-9,1994.41. A. J. Hinkle, J. R. Brockenbrough, J. T. Burg, andR. J. Bucci, Microstructurally Based PerformanceModels Relevant to Performance of Net-ShapedParts, AeroMat95, ASM Internationals 6 thAdvanced Aerospace Materials & Processes Con-ference, Anaheim, CA, May 8-11, 1995.42. C. Y. Kung and M. E. Fine, Fatigue-Crack Initia-tion and Microcrack Growth in 2024.T4 AluminumAlloys, Metallurgical. Transactions A, Vol. IOA,1977, pp. 603.610.43. M. A. Frzystupa, Development of the Microstruc-ture Based Stochastic Life Prediction Models,Final Report, Office of Naval Research ContractN00014-91-J-0199, July 1997.44. A. Heinz and W. Schelb, Improvement of FatigueStrength of AA 7050.T7451, Proceedinrs of the4th International Conference on Aluminum Alloys,1994, pp. 733.740.45. Templin, R. L., Fatigue of Aluminum, H. W. Gil-lette Memorial Lecture, Fifty-Seventh AnnualMeeting of the American Society for Testing Mater-ials, 1954.46. Butz, G. A. and Nordmark, G. E., Fatigue Resis-tance of Aluminum and its Products, NationalFarm, Construction and Industrial Machining Meet-ing, Milwaukee, WI, September 14-17, 1964.47. J. N. Yang, S. D. Manning, J. L. Rudd and R. M.Bader, Investigation of Mechanistic-Based

Equivalent Initial Flaw Size Approach, ICAF95,


17/54

International Committee on Aeronautical Fatigue,Melbourne, Australia, May 3-5, 1995.48. P. J. E. Forsyth, Acta Met., Vol. I I, 1 963, p. 703.49. C. A. Stuhhmgton and P. J. E. Folsyth, J. Instituteof Metals, Vol. 90, 1961 to 1962, p. 147.50. F. D. S. Marquis, Mechanisms of Formation ofHydrogen Porosity in 7X50 and 2X24 AluminumAlloys. Effects on Mechanical Behavior, &&Interactions in Nonferrous Metals Processing TMSAnnual Meeting 1996, Minerals, Metals & Materi-als Society (TMS), Warrendale, PA, USA, 1996,pp. 43-62.51. A. I. Hinkle, R. Skjong and G. Sigurdsson,Probability Methods for Durability and DamageTolerance Analysis, Proceedings of the 1992USAF Structural Integrity Program Conference,WL-TR-93.4080, Materials Directorate, WrightLaboratory, Air Force Materiel Command, WrightPatterson AFB, OH, Septemhcr 1993, pp. 302.340.52. R. J. Bucci, Effect of Residual Stress on FatigueCrack Growth Rate Measurement, Fracture

Mechanics: Thirteenth Confcrcnce, ASTM STP741. American Societv for Tcstine and Materials.981, pp. 2X-47.53. Metals Handbook, Volume 13, Corrosion, ASMInternational, Metals Park, Ohio, USA, 1987.54. J. A. Nicol, E. D. Seaton, G. W. Kuhlman, H. Yu,and R. Pishko, New Quenchant for Aluminum,Advanced Materials & Processes, Vol. 149. no. 4,April 1996, pp. 40%40V.55. G. W. Kuhlman and E. D. Seaton, EnhancmgMachining Performance of Aluminum ForgedProducts, Metal Heat Treating, Penton Puhlishing,Inc., Cleveland, OH, USA, January/Fchruary 1996.56. J. H. Elsner, E. P. Kvam and A. F. Grandt,Modeling and Microstructure Analysis of FatigueInitiation Life Extension by Reductions m Micro-porosity, Metallurgical and Materials TransactionsA, Vol. 28A , May 1197, pp. 1157-l 167.


18/54


19/54

Lb- Lib-(a) Reduce nitial flaw sizepopulation. (b) Reduce rack growth variability.(Quality increase) moduef/test consistency)

Figure 1:

(c) Increasecrack growth resistance.(Alloy/part design)The forging process is consistent with a classical durabilityimprovement scenario.

l In the above figure., durability is defined as the ability to prevent and controlorigination and growth of cracks to dimensions of unfavorable economicconsequence.*The Figure portrays three routes to durability performance enhancement:(a) Decreaseboth the sire and number of potentially fatigue crack initiatingmicrofeatures (e.g., reduced microporosity/better corrosion resistance).(b) Reduce variability of crack growth processes product consistency).(c) Select an alloy and/or part design to retard corrosion and growth of cracks.

l Aspects of forging microstructure and process history are well aligned with theabove durability improvement scenarios.- Forgings high degree of mechanical work can effectively heal microporosity,and/or break up deleterious remnants of the original cast structure.- Forg ings contour-following grain flow capability enables placement of theoptimal corrosion and fatigue resisting microstructure at the part surface (i.e.,where it is needed the most).- The mechanical working benefits of the forging process can be imparted inmore than one direction.


20/54

100

50PE.%0 56.-5

1

Micropore length vs. cyclic lifetime7050-T7451 thick (5.7 - 5.9 in.) plate(Long transverse, T/2 test location)I I Ill1ll~ I I I illl1.

-6.5 --1 2.0 I-- l -lPredicted bounds 70.5 dia. uniform

t 0 IS64 materialI

\29 lots. 117 tests1A 1965 material

t .(6 lots, 29 tests)I I I I IllIll \I I I,lhllII10 50 100 500 1000Cycles to failure (kilocycles)

Figure 2: Alcoa historical data shows that smooth specimen cyclic lifetimedepends on scale of the fatigue life-limiting microfeatures.


21/54

24 - hole specimen test36 * Old Qunlily--- 95% Cont. Litnils34

32

30

20

,:\

Follgue Lb (ll1ghl hours) 10 -Equivalcnl Flighl Hours

Effect of metal quality on open hole specimen fatigue performance705Gl7451 thick plate (5.7 In.), long transverse, T/2 test locationF-16 400 hr. lower wing spectrum (8000 hrs. = 1 service lifetime)Figure 3: The metal quality effect was confirmed in USAF tests employingproduction machining practices and representative loadings.


22/54

0.002 in.1980pedigreeplateMax. pore length = 0.013 n.

0.002 in.1990pedigreeplateMax. pore length = 0.006 n.

The figures show ypical crack-initiatingmicropores inked to failure of smooth atigue specimensaken rom the T/2 locationof 7050-T7451 hick (5.7 n.) plate epresentative f 1980and 1990commercial edigrees.Daring this dmeframevirtually alIaluminm mantiactarers had someeffort in place o control porosity content n thick aerospace late. Note that the material oflesserpedigree left) had he arger ailure initiating pore, and a correspondinglyower fatigue strengthcapability.Figure 4: Microporosity has been consistently linked to fatigue limitationsof thick 7050-T7451 plate.
http://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdfhttp://quality/022.pdf


23/54

23

0.6 OldO-=35 k0.4

0.20.0164 106 10

Fatigue Lifetime (cycles)Figure 5: Fatigue reliability can be improved through microstructure refinement

l The Figure shows cumulative fatigue failure probability plots for four variantsof 7050-T745 1 plate processed o contain progressively lower amounts ofmicroporosity:- Old: 1980 pedigree thick (5.7 in.) plate- Now: 1990 oediaree thick (5.7 in.) mate- Trial: sue&l pr&ess, low porosityihick (5.7 in.) plate- Thin: 1990 pedigree thin (1 in.) platel All four plate variants above were fabricated on a commercial plant sca le.l The fatigue enhancementpotential of quality-improved materials is more thanincremental.l Use of fatigue screening tests has reached specification status; thereby enablingairframers to capture the value of higher quality products.


24/54

35 958 Confidence imits 35

30 m Old material,* holes s-machined - 303 m Now material, z28 holesas-machined 2j 25 - 1 258 36 20 -.I - i *O.I15 . i 15

-. -..---. -.--.-, ....95% Confidence imitsm Oldmaterial.holesas-machined

holes e-burred

1 . . . . . . . . I . . . . ..*.I * . .

ld I& IO lo* lo* Id lo6 10Fatigue ifetime cycles) Fatigue ifetime (cycles)

.I

uIO

3/16

Open hole fatigue performance of 7050.T7451 thick plate increases with metal pedigree ;hole deburring doubles the improvement.old material: 1980 commercial production practicenow material: 1990 commercial production practice

R = 0.1 = 3OHzLT orientationT/2 ocation

Figure 6: Microstructural effects on the fatigue performance of thick productare not masked by those of geometric features or machining.


25/54
http://quality/025.pdf


26/54

1990 pedigree platePore size = 0.0044 n. Low porosity plateCOI&. part. size = 0.0019 in.Figure 8: Typical open-hole specimen fatigue fractures of 7050-T7451 thick plateshowing microporosity at the failure origin of the 1990 plate variant anda constituent particle at the origin of the low porosity variant.
http://quality/026.pdfhttp://quality/026.pdfhttp://quality/026.pdfhttp://quality/026.pdfhttp://quality/026.pdfhttp://quality/026.pdfhttp://quality/026.pdfhttp://quality/026.pdf


27/54

1990 pedigree plate Low porosity plate

Figure 9: Backscattered SEM images of constituent particles located in thehole bore of open hole fatigue specimens rom the 1990 and lowporosity 7050-T7451 hick plate variants.
http://quality/027.pdfhttp://quality/027.pdfhttp://quality/027.pdfhttp://quality/027.pdfhttp://quality/027.pdfhttp://quality/027.pdf


28/54

0.0005 in.Const. particle size = 0.0010 n.

0.001 in.

Conk particle size = 0.0009 n.

Figure 10: Open-hole specimen fractures from the thin (1 in.) 7050-T7451plate variant showing constituent particles as the failure origin.
http://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdfhttp://quality/028.pdf


29/54
http://quality/029.pdf


30/54

Figure 12: Predictive models are available that accurately link fatigue lifeto crack-originating microfeatures and vice-versa.

l Recent Alcoa modeling efforts* to relate microstructure and fatigue performance haveproduced significant tesults.- Features of microstructure which govern fatigue life have been identified and statisticallycharacterixcd for several material variants and specimen types.- Links between ife-governing microfeature populations and fatigue performance havebeen defined, modeled and validated.- Size distributions of crack initiating microstructura l features can be measured directlyfrom fracture surface examinations, back-calculated from lifetime data, or scaled fromrandom plane microstructure.- The effects of the initial inhomogeneity size, shape, and location distributions onspecimen fatigue hfe distributions have been demonstrated and analytically predicted.- Characteristic microstructural feature distributions derived from basic coupon testshas been applied to predict structural element petfonnancc.l Penetration of corrosive attack found in moderate to severely corroded metallic aircraftstructure is comparable to, and often exceeds, he dimensions of fatigue. lie-limitingmicrofeatures commonly associated with failure in pristine element tests. (typically, 0.5 to20 mils in the latter).

* Support f thesemodding effortsby theOffice of NavalFtesearcbONRConuaaNooO14-91-C-0128)is gratefullyacknowledged.


31/54

Y40 0 0 00 r- 0 0 0L-J0 0 0 0 00 0 0 0 0By assembly By structural detail(e.g. fastener holes)

Figure 13: A coupon to component transfer model has beendevised based on life-limiting feature population.


32/54

L.iie-limitingmtcrofeature lstributlons ithreematerial ariants. urn0.6 , I h&rofeature distributioncanbe related opotential or multi-site damage.Calculated o. flaw exceedancesper loo0 fastener oles*

Ftaw one ifetime** two liietimea**ii?, A B C A B C0.03 196 3 0 850 850 440.05 2 0 0 633 499 10.1 0 0 0 195 3 00.2 0 0 0 3 0 0* calculatedfterManning nd ang** Fighter lower wing specuum

Figure 14: Probabilistic fracture mechanics models can be used to study theeffect of microfeature distribution on structural performance+

l Consider the example above from 7050 plate. Three different equivalent initialflaw size (EIFS) distributions (linked to microstructure of three 7050 processvariants) are shown in the figure on the left. For each of these crack startingfeature distributions, the expected number of cracks exceeding a given size per1000 fastener holes was calculated for one and two lifetimes of anticipated serviceloadings. It can be seen that the distribution linked to process variant C has thelowest probability of developing multi-site fatigue damage.l The lack of EIFS distribution data for representative materials and structurespresently constitutes a major hurdle to broader application of these tools in des ignand risk assessment rocesses.


33/54

I_ 37*22n n n n nLr -z ===-===-- -ii-- ----- -=+-l---i11 u u uAREA: 260.55 SQ.IN I u

T.001p 37.221 R .25 1 n

cr-=--I-1.76I--_u u70.03 SQ.lN - 4.76

Figure 15: Closed Die Forging vs. Thick Plate Raw Material Product Comparisonfor a Typical Machined Web-Rib Airframe Component.


34/54

4020

010' 105 10

- lPId6-o- 2Plal.a* 3P ate- 4Plal.s- sp ateb 5.7Plat.- 6 Low Porosity

+ DidMIFail

Cycles to FailureCumulative Fatigue Failure Distributions7050-l7451 Plate (1 - 6 in. thk.)(Max. stress = 40 krl, R = 0.1, kt = 1. long-transverse, T/2 test location)

Figure 16: Plate tatigue performance is affected by product thickness andmanufacturing pedigree.l Engineers have known for some time that microporosity can shorten life of cyclicallyloaded components by accelerating the initiation of fatigue cracks.l In thick plate, microporosity persists through to the final product where deformationduring rolling is often insufficient to effectively break-up and/or heal porosity present inthe stming ingot.l Increased mechanical work in rolling to thinner plate gages decreasesboth themicroporosity sire and volume fraction, and hence fatigue resistance s improved overthat of thick plate.l Thick plate fatigue improvement potential has been demonstrated through control ofmanufacturing processes o curb microporosity.l Because of the loss of inherent crack stoppers n consolidated parts, new cost reductioninitiatives seeking to replace conventional assemblieswith fewer, thick parts mustconsider control of microporosity as a fatigue crack preventive measure.


35/54

Cumulative Failure Percentage


36/54

10 lo5 1 o= 10Cyclee to Failure

Cumulative Fatigue Failure Distributions7050-T74XX Thick Products(Max. stress I 40 ksi, R = 0.1, kt I 1, long-transverse, T/2 test location)Figure 18: Product form (forging or plate) affects material fatigue performance.

Dieforging fatigue resistance is superior to that of thick plate product needed to machine the same part.

l The fatigue resistance of heavy section die-forgings is typically much greater than that ofthick plate product needed to machine the same part.l Hand forging fatigue resistance s generally equal to or better than that of standard (1990pedigree) thick plate, but somewhat inferior to that of the process mproved low porosityplate.

l Hand forging fatigue performance is also limited by microporosity, and lie plate,enhancement should also be possible through implementation of improved processingmethods for control of microporosity.l In the > 6 product thickness range hand forging fatigue resistance generally remainsstable, whereas plate fatigue resistance generally degradeswith thickness increase.


37/54

The figures show partially healedmicropores ound on fracture faces ofshort-lived specimensaken rom a 6 in. thick 7050-T7452 hand forging.Figure 19: Microporosity is sometimes found to be the failure-initiating microfeaturein tests of smooth fatigue specimens aken from thick hand forgings.
http://quality/037.pdfhttp://quality/037.pdfhttp://quality/037.pdfhttp://quality/037.pdfhttp://quality/037.pdfhttp://quality/037.pdfhttp://quality/037.pdf


38/54

38

cum.f&Jrepercent

60

+---c-106 106 IO7 108

0 c .-.-. _..~1 plate3 PlateDidIlOIFail

Web Area (0.65-1.36)

Cycles to FailureCumulative Fatigue Failure Distributions705@T74XX Die Forging and Plate Products(Max. stress = 45 ksi. R = 0.1, kt = 1. long-transverse, T/2 test location)

Figure 20: Cyclic lifetimes of specimens from heavy-worked die-forgingsexceed those of specimens from thin plate.


39/54

me&4Cycles to FallunEffect of process route on cumulative fatigue failuredistributions of 7050-T745X thick products(Max. stress = 40 ksi, R = 0.1, kt = 1, short-transverse, T/Z test location)

Figure 21: Process route affects forged part performance.More sophisticatedprocessmute B increasesatigue performance.

l Forging process B employs a greater number of manufacturing steps to controlmicrostructure (viz., m icroporosity), and thereby allow higher fatigue resistance to beattained.l Die-forgings, and to lesser degree hand forgings, offer a variety of alternative paths toarrive at the same shape part shape. Optimal use of the forgings will result when userand vendor blend expertise and tune their respective manufacturing processes orequirements of the end-part. Such relationships may be necessary o offset thenegative implications of forgings higher initial cost.l Much reCentprogress has been made in developing modeling linkages amongprocessing, microstructure. and performance. These models are being used to challengeconventional airframer material selection processes o consider the benefit potential ofquality-improved materials to gain an overall end component cost/performanceadvantage.


40/54

Comparison of 70501T74XX products open hole fatigue performancefor thick plate (5.7 in.) and die forging (6.0 in.) and thin (1.0 in.) plate(LT orientation, T/2 test location , R = 0.1)34.0 I

32.0 - \ un \

I I IA 6.0 Me Forgings (0.66.1.36 Web) _

- 95% conlldenca llmitr (6.0)0 5.7 PIat* Hohs Daburnd

30.0 -af 26.0 -isf 26.0 :Eg 24.0 -z3 22.0 ;

r

- - - - 65% confidence llmlts (6.7) i0 1.0 PI610_... 06% confidence llmltn (1.0)

0.167 0

1.00

i

00.126 thick

20.0 :16.0 10'

-I- Did not fall -

I LLlo6 logLife (cycles)

T.00

1

WM-lo-14 PPB

Figure 22: Smooth fatigue rankings prevail in the presence of a design detail.Open hole fatigue performance of I in. plate and die forgings are superior to that of thick plate.


41/54


42/54

Const.-load-amplitude FCG data for thick 7050-174 closed die forgings(L-T orientation, R = 0.33, high humidity air)AK, MPadm

1

1

1

1

-5 _0 i

0 -6

.7-0 3

.60 p

1 5 10 30AK, ksih.

1 o.6al1o-7 'j,9E1o-83

1 o.g

1OO

Figure 24: Fatigue crack growth rate properties of 7050 forging and plate are similar.


43/54

FCGR data from thick, incompletely stress relieved 7050-174X hand forgings(S-L orientation, R = 0.33, high humidity air)AK, MPadm5 10 20

yl o.5

yl o.6210. gPE

Fl o-E g-ax7yl 0-O

i 1 OOIO.01

1 5 10 20AK, ksih.

Figure 25: The industrys forging FCGR data base s tainted byresidual stress induced artifacts.The orging residual stress effect s magnified in coupon tests.


44/54

Closure corrected FCGR data from two incompletelystress relieved 7050-T7452 hand forging (6 in. thk.)(S-L orientation, R = 0.33, high humidity air)AK,Mpa(m)ln5 10 20I 0 Spe&llen1

I 9peohn.n2

f,E lo*Ba= IO"

:lO-'"lo+ I I-

1AK,

l&n)~@ 10 20

Figure 26: A corrective practice has been devised to remove residual stressbias from FCGR data.FCGR data from residual stress containing specimens are collapsed into a single curvewhen the crack closure based corrective p ractice is applied.


45/54

Closure based FCGR data for various th icknesses of 7050-l7451 plate(S-L orientation, R = 0.33, high humidity air)AK, MPadm5 10

1 oe5

5 10 30AK, k&n.

Figure 27: Closure corrected FCGR data from 7050 plate and non-stress relievedforging are similar.Thus, intrinsic FCG res istances of plate and forgings can be considered equivalent, andplate andforging data can be used interchangeably for purpose o f FCG life estimation.


46/54

Coupon test S-T T-L L-T

I Plate grainflow axes Cyclic stress intensity, AKFCGR properties formajor plate grain flowsI 1 I11 '6 / 1 1ocation 1 Plate 1Foraina 1

Component

Figure 28: The benefit of contour-following grain flowon forged part life can be estimated fromplate data.


47/54


48/54

48

Inilbl aba dbtlbutbn

20i8 --16 --

14 --12 --10 --

m -WE3 lOE8 lOE7 lOE9

cycle0 to tahrePredicted open hole S-N response of forging vs. plate

Figure 30: The fatigue benefit po

comparação forjado x chapa

Documents

Transcript of comparação forjado x chapa