General Disclaimer One or more of the Following Statements ...are reviewed and analyzed. Thin-sheet Charpy and Izod Impact tests and standard full-size Charpy impact tests were conducted

General Disclaimer

One or more of the Following Statements may affect this Document

This document has been reproduced from the best copy furnished by the

organizational source. It is being released in the interest of making available as

much information as possible.

This document may contain data, which exceeds the sheet parameters. It was

furnished in this condition by the organizational source and is the best copy

available.

This document may contain tone-on-tone or color graphs, charts and/or pictures,

which have been reproduced in black and white.

This document is paginated as submitted by the original source.

Portions of this document are not fully legible due to the historical nature of some

of the material. However, it is the best reproduction available from the original

submission.

Produced by the NASA Center for Aerospace Information (CASI)

https://ntrs.nasa.gov/search.jsp?R=19760011148 2020-05-16T13:15:29+00:00Z

w ^ ^

P • ^

7

{

3{

d

NA&4'",4 TECHNICAL NASA TM X-71875MEMORANDUM

u,1

X (NASA-TM-X-71875) IMTROVED IMPACT-RESISTANT N76-18236BORON-ALUMINUM COMPOSITES FOR USE AS TURBINE

:I*.-'

ENGINE FAN BLRDES (NASA) 29 p HC $4.004 CSCL 11 D Unclas

r G3/24 18502

IMPROVED IMPACT-RESISTANT BORON-ALUMINUMCOMPOSITES FOR USE AS TURBINE ENGINE FAN BLADES

by David L. McDanels and Robert A. SignorelliLewis Research CenterCleveland, Ohio 44135

TECHNICAL PAPER presented atSymposium on Failure Modes in Composites

` sponsored by the American Institute of Mining,Metallurgical, and Petroleum EngineersLas Vegas, Nevada, February-23-24, 1976

m,

t

w

t,

s

z f

114PROVED IMPACT-RESISTANT BORON/ALUMI NU14

COMPOSITES FOR USE AS TURBINE ENGINE FAN BLADES

by David L. McDanels and Robert A. Signorelli

NASA-Lewis Research CenterCleveland, Ohio

ABSTRACT

{

Efforts to improve the impact resistance of B/Alare reviewed and analyzed. Thin-sheet Charpy and IzodImpact tests and standard full-size Charpy impact testswere conducted on unidirectional and angleply compos-ites containing 4, 5.6 and 8 mil boron in 1100 0 2024,5052 and 6061 Al matrices. Impact failure modes ofB/A1 are proposed i n an attempt to describe themechanisms involved and to provide insight formaximizing Impact resistance.

The Impact strength of B/A1 was significantlyincreased by proper selection of materials andprocessing. The use of more ductile matrices (1100 Al)and larger diameter (8 mil) boron fibers gave thehighest impact strengths by allowing matrix sheardeformation and multiplefiber breakage.

Pendulum impact test results of improved B/A1 werehigher than notched titanium and appear to be highgpqugh to give sufficient foreign obje-ct, damage protec-t 'ion- to warrant consideration of B/Al for applicationto fan blades in aircraft gas turbine engines.

'S INTRODUCTION

Studies by 14ASA and the Ai r Force have shown theadvantages of using composites as rotating fan andcompressor blades in turbine engines. Composites offer

f,

ORIG^^ POF p^R QA OE i4r

P

sa

TT _,-. air

f s

rt

lighter weight, lower cost, and hi gher specificstreng th and stiffness, resulting in improved engineperformance and lower direct operating costs.

Most prior materials development has been directedtoward using high specific strength and stiffness com-posites for airframe structures. High mechanical prop-erties are most important for these applications andlittle attention has been given to impact resistance.

However for rotatin g fan and compressor blades inaircraft engine applications, impact and foreign objectdamage ( FOP) resistance become as important to opera-tional performance as strength and stiffness. Ref. 1defined a foreign object debris spectrum as small-bodyand large-body damage. Small-body damage includes hardobjects such as sand, rocks, rivets, and ice balls.Large-body FOD is caused by hard bodies such as iceslabs, and soft bodies such as birds. Localized damagefrom small-body impact can result in minor reductionsin fatigue strength, whi le large-body impact may causenomplete airfoil separation requiring a reduction inengine speed or complete shutdown.

Collisions with birds are a major flight safetyhazard encountered in aircraft operation. Mostcollisions' occur with birds ranging in weight from 4ounce starlings to 4 pound ducks. During the 1967-69period, 35 of all aircraft accidents were attributableto bi rd strikes (ref. 2). About 52 of the birdpopulation (fib;. 1) occurs at altitudes less than 500feet, endangering take-off and landing operations.Take-off conditions are the most severe since theengine is required to operate at full power and powerreduction or loss could be catastrophic.

Lack of FOD resistance has been a major obstacleto the use of composites as fan blades in aircraft en-g ines. Composi te blades have shown considerable prom- -Ise in preliminary testing, but in full-stage enginetests, the results have been less than satisfactory.The results indicate that composite blades must have

' additional impact resistance to become competitive withconventional titanium and stainless steel blades. inadd i tion, root attachment methods used for the bladeshave caused fiber breakage during fabrication,

` - resulting in premature failure during en g ine testing.

To overcome these problems, NASA-Lewis has con-ducted in-house and contractual work to improve impact'resistance of both polymer and metal matrix- composites

2 -^

f

k

i

for fan blade applications. The objective of thisreport is to review the programs supporting the Impactimprovement of B/Al composites and to analyze some ofthe factors that can increase the impact resistance ofmetal matrix composites. The results and analysis ofthe NASA-LeRG in-house programs are presented ingreater detail in refs. 3-4, and the contract resultsIn ref. 5. Tensile tests and Impact tests onthin-sheet and full -size specimens were conducted todetermine the effect of processing variables, matrices,fiber diameters., and anglepl ies on the impactresistance of B/Al composites, Impact failure modesare proposed and are related to the results obtained.

MATERIALS AND PROCEDURE

i+{aterials Selection

Commercially produced boron fiber, of 0,10 mm (4mil), 0.14 mm (5.G mil), and 0420 mm (8 m1l) diameter,was used for composites in this investigation. becauseof the standard nomenclature used in the aerospaceindustry, boron fiber diameter will be referred to inmils, rather than in SI units, throughout this report,

Aluminum alloy matrices, 1100, 2024, 5052, and1 6061, were selected to cover a range of 'impact

strengths and ductilities,

Specimen Preparation

All B/A) panels for the in-house study nominallycontained 48 volume percent boron and were made by

f press diffusion bonding of fiber layups between matrixfoils. The first series of panels, consisting of 8-plyunidirectional -8 mil x/1100 Al 'composites, were used todetermine the effect of fabrication temperature onImpact properties. These panels were bonded attemperatures from 714 K (825 F) to 783 K (950 F).

After selection of a standard fabrication condi.°t i on of 755` K ( 900 F) for 0.5 hour at 34 HPa (5 ks i ),another series of 1100 Al matrix panels was fabricated,In addition 2024 Al panels were fabricated at 774 K(935 F) and panels with 6061 Al and 5052 Al werefabricated at 805 K (965 F). These panels were also

z; bonded at 34 14Pa (5 ksi) for 0,5 hour, An g 1ep 1'y layupswere symmetrical from the center, The '8-ply panels

i were used for tensile' and thin-sheet impact, tests,` Panels for full-size Charpy impact' tests were 40-ply

,ifor 8 mil boron ., 60 -plat for 5,6 miland 80-ply for 4

k

i

rill 14 The full-size Charpy specimens were surfaceground to AST14 specifications and a 45-degree notch wascut into one face.

Specimen Geometry

Because of the anistropic properties ofcomposites, specimen geometry must be uniquely definedin terms of fiber direction, pressing direction andnotch location, These geometries are shown in fig, 2.The LT, TL, and TT geometries were defined in refs,6-7o The LT geometry was further' defined in ref. 5 asLT, where the testing direction was in a plane normalto the pressing direction, and LT(s), where the notchwas on a side parallel to the pressing direction.Tests were conducted on specimens with LT, LT(s), andTT geometries for the studies reported In this paper,

Impact Tests

Three types of pendulum im pact tests were con-ducted: unnotched thin-sheet Izod, unnotched thin-sheetCharpy, and notched full-size Charpy . Thin-sheet testswere conducted because they are more economical Interms of material and machining costs and serve as aconvenient screening tool. The cantilever mounting ofthe izod test tends to simulate the behavior of amodern_, thin-airfoil fan blade in engine operation.Thin-sheet Charpy tests provided an indication of theunrestrained behavior of the material Full-sizeCharpy tests provided a comparison of standardspecimens with literature values of other materials.

RESULTS AND DISCUSSION

Tensile Test Results

L it d'1 t '1' t 4-k of 8 1 B/1100 Al%J"6 u ina ens e s reng mimatrix composites decreased with increasing angleply

` (fig, 3)a Longitudinal stress-strain curves, fig, 4,were plotted until the load started dropping, asindicated by the arrows. Unidirectional specimensshowed linear behavior to failure, With increasingangleply, noelineari y and strain to maximum loadIncreased. At failure, the specimens started toseparate along the angleply axes.

Transverse stress-strain curves, fig, 5, show thatx _ angleplying increased the strain to failure of the 1100'

Al matrix composites and Increased the transversestrength slightly,

.. 4

a v

r

Itilpect Test Results

Fig b compares the area-cor.1pensated LT i mpactstrength of undirectional 1100 Al compos ites fur threedifferent moron fiber d;aweters , The area under thenotch was used for area compensation of standard full-size Charpy specir•lens? The thin--sheet specimens wereunnotched and the entire cross section was used furarea cor,ipensatiun, For each type of test,, the area-compensated impact strength increased with increasingfiber diameter., The values for full-size Charpy testsof 8 tall boron specimens are shown as a band becausethe e mil b unidirectional panels used for the in-houset(:sts were inadequatel y bonded and gave_ excessively lowvalues There frre the lowe r bound represents ' ex t<rapo-l of i olds from in-house angl ep l y test results The upperbound represents Impact values from ref,, a. In ei thercase, the increase in impact stren g th from 5,6 to 8 milboron specimens is considerably greater with thefull-size Charil y tests than with the thin-sheet tests,

Tile area-compensated full-size Charpy Impactstrength was inuch higher than that of thin-sheet s peci-wens, Properly bonded full-size Charpy and unidirec-tiunal and low an g lep'l y thin-sheet speciisiens failed byfracture of all fibers in the cross section, Frithmatrix plastic shear priur to fiber failure, Full-sizeCharpy specimens had snore shear than thin-sheetspecimens. higher angleply specimens underwent bendingdistortion but were pushed through the grips at; lowImpact energies with winimum fiber breakage;;

The difference in area-cor,lpensated impact strengthvalues fur thin=sheet and full-size impact tests isrelated to their thickness and failure mechanism. Refs,8-10 reported a transition in fracture and delaminationbehavior at a thickness of 0.25 cm (0.1 inch) Below

ithis thickness, plane stress;_ conditions applied anddelaraination stresses were very high, Fiber/matrixbond fai lure occurred due to shear stress concentrationat the notch tips Above this thickness, plane strain'conditions applied where transverse tensile stresses atthe notch tip caused fiber/watrix bond failure at loverstresses and the stress to cause de'lamination remainedconstant. In both cases, after the notched section deglaminated, the remaining section was notch-insensitiveand failed as if a notch had not been present (ref, 8)

These results indicate that thin-sheet impacttests can be used as a screening tool to rank impactbehavior of var ious B/Al composites, but the quantita-

()RIGINAE. OF POOR

UAGE ^gQ ,1 LZIl1s

S

t

I

S

C'Cl 4

xF

3

Live results of one type of test cannot be extrapolatedto another. it shoul d be noted that the comparison ofarea-compen sated results from full-size and thin-sheetspecimens appeared to be vali d qualitatively despitethe fact that the thin- sheet specimens were below thetransition thickness of 0,1 inch while the full-sizeCharpy specimens were above. This probably influencedthe inability to extrapolate quantitative values fromone test to another. The rankings were consistent fortests on different matrices, fiber diameters, andanglepl ie_s where failure ocurred, in anglepi ies wherethe thin-sheet specimens deformed, but did not fail,thin-sheet results could not be used for accurateranking purposes. The indicated impact strengths wereactuall y a measure of; 1) impact strength, if thewaterial were strong enough or brittle enou gh not todeform excessivel y , or 'l) bending stress, if thematerial was pushed through the holders without fiberfracture, or 3) a combination of the two, where thematerial partially deformed and partially failed.

Factors Influencing improved Impact Behaviorof Boron/Al urninui Composites

One of the problemsinherent in the evaluation ofcomposite toughness is that a variety of testingmethods- have been used. Interpretation of results aredifferent depending upon whether notched tensile testsor bending/impact tests are conducted. The ends arerigidly restrained in tensile tests, while in slog bendor Impact tests, both ends may be free (Chirpy) or oneend i.iay be cl amped Ozod), Although strength in bendingshould be comparable to strength in tension, the strainbehavior is different. Therefore, interpretation ofresults and prediction of behavior should be approachedwith caution when comparingfracture toughness, work offracture or "impact strength results from differenttypes of tests.

iJotched Charpy and Izod impact tests are accepted`as convenient methods of determining the susceptibilityof a material to brittle' fracture at high strain rates.

-Although data from these `tests have been used with somesuccess, the approach has been largely empirical (ref.'11)^ For homogeneous materials, the effects of notch

x geor^ietry and elastic and plastic deformation under plane stress and `plane strain ` conditions at both thenotch region and throughout the specimen are verycomplex o The stress state and toughness behavior ofcomposites are even more complex because of the

? divergent properties of the two constituents,

r:^.xG

ORIGnvAU PAGEOF POOR QUA'$

t

s £

►

ileasurement ctfraures UELQ=O _ Ref, 12 states thattwo different concepts can be used to measure fractureenergy, One involves measurer.Ient of the total energyintroduced into a specir,len during fracture, averagedover the entire fracture process.. This categoryIncludes work of fracture and Charpy/izod impacttesting, The other involves ► iieasurernent of the initialrate of strain-energy release at failure and includesfracture mechanics analyses pertaining to initiation offracture Results on carbon fiber reinforced glass(ref, 1 1) shoared that work of fracture, which incl uded Ifiber failure and fiber pull -out, was much larger thanthe energy required to initiate fracture,

iki) empiri ca l relation to predict imp act propertiesof composites was presented in refs, G-7, Goodagreehlent eras reported In the prediction that Impactstrength of G/Al may be increased by increasing thetensi le st reng th, volur,ie percentage, and diameter ofthe fiber and by decreasing the shear strength of theiaatr1x, This relation may be val id for predicting;cneral trends, but is probably not val id for exactcalculation, The apparent agreement noted in refs. E-7wa y be coincidental n

Results obtained in the NASA- Le RC programs show! that the impact energy of B/Al composites also depends

upon other factors, related to fabrication conditionsi and -failure mechanisms. This dependence was predi cted

in ref„ la where Impact ener gy density (strain energydivided by volume) tivas shorn to be influenced by acorre lation coefficient, which is a complex function ofconstituent propert ies base d upon fabrication hi story.

i?^1 ^"SZCI 52 Sd M 110" Ju In"" eners:v absor12tIo-nRefs 14 and 15 reported that work of fracture of

composi tes is influenced by the stren gth and fractureuehavlor of the fiber, the matrix, and the interfacebetween the tiro, Contriuutions to energy absorption byeach are interrelated and can l imit or enhance thecuntributions of the others

Table l summarizes the relation of fracture modeto hopact energy absorpt ion possible in [3/A1 , Thelowest ei ► e ray absorption would by from cleavage

4 failures, Although not encountered in this program,cleavage failure could occur in overbonded compositeswhere interfacial reaction has forced the fiber to lose

i01? G

PA 4EOF pppl? @UA y

^' 7

its Identity and failure would occur in a manner sit a i-lar to brittle hortioseneous materials, A planar frac-ture would have slightly higher energy absorption, Inplanar fractures, energy absorption would be prii- iiarilycontrolled by the f lber fracture energy.. with no matrixcontribution,, Delarilination or f iner pull-out failureswould have inedium Impact energy absorption.. I n de I am-Ination, energy is absorbed by surface energy releaseupondelamination of the B/Al or AI/Al interfaces.With fiber pull-outo energy Is absorbed by frictionalsliding and plastic shear at the InterfaceA Failure bymatrix shear with single fiberfailures gives highenergy absorption because each component makes acontribution to the energy absorbed by the composite.The fiber contribution comes from fiber fractureenergy, while the matrix and the Interface contribu-tions are by shear displacement energy. Matrix shearwith multiple fiber breakage gives the highest Impactenergy absorption. In this case the fiber contributesadditional energy absorption because of multipleb re a kag,-e and the matrix contribution is increasedbecause of the additional plastic shear allowed*

While the table indicates the relation of failurewode to hilpact energy absorption, it does not indicatehow the toughness of cortiposites can be Improved. Inthis paper, the materials and processing variables thatcan increase composite toughness by exploitation ofthese fracture modes will be discussed.

;Ufec t Q—f ±a^_r Lao_LLQa t er1luaX_Q1ur&- - Impact resistanceof B/Al can be increased by the use of fabricationterilperatures that allow adequate bonding (to preventdelamination and to make failure dependent upon fiberfracture energy) to obtain properties required for agiven application. At the same time, the temperaturemust be low enough to prevent excessive aluminum borideformation (so that the fibers can exhibit maximumstrain to failure),

Area-compensated Izod Impact strength is plottedin fig, 7 for thin-sheet specimens bonded for 045 hourat various temperatures. Two curves are plotted onthis figure: one tor delamination failures and theother for fibrous failures. For delamination failures,the impact strength increased with increasingtemperature, due to improved bonding with temperature.Fibrous failures did not occur at lower bondingI-c%rn arni-"rac Whara fikro"c fil"rc nr,-iirrad theImpact strength decreased

failures temperature,

ORIGINAU PAGE LS

A

F

SPecimens fabricated at lower temperatures failedUy delamination at low area-compensated lzod impact en-ergies. Bonding was not adequate at these temperaturesto allow the corilposites to attain theli e, full impactstrength,, Tile fiber/matrix Interface was weak and somespecimens even delarilinated upon machining prior totes t 1 11 g 0

At higher bonding temperatures, the area-compen-sated thin-sheet Izod impact strength Increased$ Withadequate bonding, the stress to cause delamination atthe fiber/matrix interface Increased and the watrixcould undergo sufficient shear deformation to fracturethe fibers. Thus for max Imum impact resistance,, thefailure mechanism changed from being Interface controlled delamination to being fiber fracture controlledo

The maximum area-compensated Impact strength forB/1100 Al was in the 741-755 K (875-900 F) range in theNASA-LeRG In ,-house prograin. Ref. 5 reported thatinaxiiiium Impact properties were obtained using theirfabrication cycle at 727 K (850 F). Thus there isprobably a range over which maxiiiium impact resistancecan be obtalnedo This range would be dependent uponthe complete fabrication cycle used, and upon the foilsurface condition and amount of deformation present.

After f-i-abrication at temperatures in excess of 783I's (950 F), the impact strength _ would probably dropfurther, due to property degradation from fiber/matrixinterfacial reaction. The formation of a thin brittlephase layer at the Interface reduced the strain cap-bility of the fiber, thus reducing tensile and impactstrength. Although Impact data were not obtained fromspecimens bonded above 783 K (950 F), degradation hasbeer) reported by others after processing at highertemperatures,, The fatigue limit of B/6061 Al com pos-ites was reduced by increasing the bonding temperature(ref. 16). Ref,, 17 reported-a 2U", increase In full-sizeCharpy Impact strength of Borsic/6061 Al composites to9.4 joules (74 ft-lbs) by reducing the bondingtemperature from 838 K ( 1050 F) to 723 K ( 8 42 F) .

9 =rU . - The purpose of the matrix is toflai= Q-f -provide sufficient ductility to permit the fibers toattain their full strength during the impact processaWith sufficientmatrix ductility, the fibers can morenearly approach their full strain capability O andfailure can occur in an optimum manner where the matrix

A t he %C I k 1fu ll1 * 1, I

fractureWZ I MO '^u C3 U Lovlj L I UL t U" ;.0 Lit: energy.

In this program, U tall di arieter boron fibers wereused to reinforce four aluminum alloy matriceso, 1100,5052, OU1. and 2024 These alloys representeddifferent coolbInations of strength and ductilityoLiterature data cited for those alloys serve only as anIndication of anticipated behavior in composites sincethe stress-strain behavior Is chan ged by restraint bythe fibers,; Shear strength becomes an importantcriterion only If the shear strength of the matrix Istotter than the shear strength of the fiber/matrixInterface,, This was demonstrated In ref, 10, whichshowed that the fracture toughness of Borsic/1100 Alwas independent of B/Al interfacial bond strength.

Ref, 19 proposed that for matrices where thefailure strain is higher than that of the fibers, acrack will propagate by sequential failure of thefibers, followed by failure of ti(; rtiatrix along a linejoining adjacent fiber breaks. If there Is a flaw-dependent 11 angth- s t reng th I ref, 2 0) ef fect, ,,,here thefibers break at different stresses, fiber fracturesWill riot be alined and matrix shearing will occurbetween fiber falluresR This situation Is shown sche-hiatically In fig. 8-a,, Analytical prediction of workof fracture for this case is difficult because of problems in determining the total area undergoing shear.If the strengths of the fibers are uniform and they donot have flaws distributed along their length, thefracture will be nearly planar and the crack will notbe deflected from a path directly across the speclmen;,This wouldbe the case for a plastically deformingfiber such as ductile tungsten wire, Under theseconditions, no fiber pull-out would occur and work offracture would be determined by contributions fromplastic deformation of the components, In the case ofbrittle fibers,, such as carbon or boron, fracture Isinitiated by sequential failure of the brittle fiberson a plane normal to the tensile axis., Ref, 19 statesthat fracture of brittle fibers should absorb littleenergy and that the plastic deformation of matrixbridges connecting fiber lengths on either side of theincipient fracture will determine the work of fracture.

For matrices where the failure strain Is lowerthan that of the fibers, failure will be Initiated bythe growth of a crack in the matrix (ref, 19). Thiscrack will tend to be planar, and unbroken fibers willbe left bridging the crack. These fibers will falleventually at weak points adjacent to the plane of thematrix crack. The matrix fracture surface will be

10

r

smooth with some surface depressions and projectingpulled-out fibers, Thi s s ituation is sho rn in fig.U-b. in this ca g e, worst of fracture can be predictedusing the analysis of ref, 20,

Refs 14 and 19-22 reported that maximum work offracture occurs with discontinuous fiber composites.When a crack passes through a composite, fibers shorterthan the critical length are pulled out from thewatrix, rat=her than broken» Fibers of the criticallength have a maximum distance of pull-out, Fiberslonger than the critical length will fail in tension,normally at a lower work of fractu.re,. Work of fractureis thus a combination of the work needed to debond thefibers from the matrix and the work done in Pullin,-, thefibers out of the matrix. However, it should beemphasized that this occurs primarily in the case wherethe rolatrix is more brittle than the fibers (ref, 19).

For the case where the fiber is ductil e and therilatrix is very brittle, fracture would be initiated inthe brittle matrix. Multiple cracking of the matrixcould occur because deformation is not limited to thep lane of final fracture,

Results of this program follow the behavioroutlined above, Thin-sheet izodand Charpy, as well asfull-size Charpy impact strength of Q/Al was increasedusin g more ductile and weaker matrices., Composites jwith 1100 Al matrices hadsignificantly higher impact 1stren gths than those with other matrices, Compositeswith stronger and less ductile matrices had the lowestImpact strengthso Similar results were reported inref- 5. The fracture surface became more gagged andIrregular with increasing impact strength, and fiber/matrix projection zones of fibers connected by bondedmatrix were projecting out of the fracture surface.Fig. 9 shows comparisons of fracture surfaces for B/Alcomposites observed in ref. 5, For 54 mil B/1100 Alcomposites (figs 9-a), some bare fiber pull-out can beseen at the Cops of some of the projection zones, bu tthe general ,jaggedness and projection zone formation isapparent. Fi g, 9-jb shows that the pr oject ion zoneeffect is more pronounced with the higher impactstrength B mi l boron composites, Fig, 9-c s hows thebrittle, planar fracture surface of a lower impactstrength 5052 Al matrix composite with no fiber/matrix.projection zones present

Fig. 10 shows failed full-size Charpy specimens,The low-energy fracture of the 5.6 mil B/5052 Al

11 ^

e

a

l.^

7

cort► poslte (fi g . 10-a) was p lanar with no matrix shear.Restraint by the boron fibers reduced matrix ductili tybelow its unreinforced value The 2024, 5052, and 6061Al Matrix composites acted In the matrix-less-ductile-than-fibers case of ref,, 10, The ductility of the 1100Al matrix was sufficiently high to be more ductile thanthe fibers. The higher-energy 5 b mil 13/1100 Alcomposite ffig,, 10 -b) shows a ,lagged fracture surfacewith a large amount of shear deformation, F14¢ 11,shows that the shear displacement at the ends of tailedLT t• ull-size Charpy specimens from ref. 5 increaseslinearly with increasing Impact strength Thisdeformation increased the h^1pact strength of thecompos ite In two ways. Firs-t, additional ener gy wasabsorbed through ,Multiple breakage of the fibers.Second, the matrix absorbed more energy throughadditional shear after the Init ia l fiber fail ures,

In high impact strength (3/1100 Al composites, theinatrix sheared during pendulum impact and the fibersfailed In tensir,,a,, With the additional matri x shearallowed by the ^.ui fAle 1100 Al matrix, the tensilestresses in that intact portions of the broken fibersconti}rued to Increase and failed the fibers a;ain.Composites with 8 rni 1 boron showed more rnatr i x shear

^

ductility and multiple fiber breaha geq Fig. 12 shown afailed 8 isil1 13/1100 Al thin-sheet Izod specimen. Theouter fibers have radial cracks in the fracture regionat fairly regular distances along the fiber length.This Indicates that multiple fiber breaka ge occurredpr ior to and during failure. This mul tiple fiberbreai;-,ue was localized in the fracture region.

Gf.^ t _QJ figs c i aWg r. - Area-compensated LT Impactstrengths of 110(1 Al tiiatrix composites with va -iousfiber diar; ►eters are shown. In fig, G for three types ofImpact tests,, These resul ts Indicate that the Impactstrength of 6/Al increased with increasing fiber diam-eter,, Ref. 5 also reported that the impact strength ofB/1100 Al was higher using 8 mil boron than with 5.6iii il^ Limited data in refs. G and 17 showed similartrends, lVork of fracture for copper matrix compositeswith brittle, recrystal lized tungsten wires alsoIncreased with increasing fiber diameter (ref. 10).

For a given fiber content, increasing fiber diam -eter decreases the total surface-to-volume ratio of thefibers within the composite. Increasing the 'diameterfrom 4 to 5 t G rail' l s, or from 5.6 to a mils doubles thecross-sectional area of a single fiber, but onlyincrease's the shear area by 401,, The shear stress

12

would be higher at a given tensile load 4 allowing aductile matrix and/or fiber/matrix interface to yieldand shear prior to composite fractured Shear isdesirable if the matrix has sufficient ductility toallow plastic shear without premature crack initiationprior to fracture,,

Interfiber distance must be great enough to allowthe matrix to exhibit its full ductility and to absorbImpact energy by shear deformation. The increase Ineffective fiber diameter caused by restraint of thematrix by the fibers (ref, 23) reduces the distancebetween adjacent fibers for accommodating shear dis-placement. This effect decreases with Increasing fiberdiai,ilete. r since Interfiber distances are correspondinglylarger for a given fiber content. Specimens with 4 milboron displayed little shear during fracture and had,ie lowest Impact strengths " No multiple fractureswere observed and the ductility of the 1100 Al matrixwas minimal. The Increase In effective fiber diameterreduced the already small interfiber distance evenftirther and the matrlx could not act In a ductilemannero

Increasing the boron diameter to 5.6 milsincreased the Interfiber spacing. These specimensexhibited Increased fracture ductility and Impactstrength, In this case the Interfiber spacing wassufficient to allow some shear and multiple breakage.

Comparison of figsn 9-a and 9-b shows that the 8ml l boron specimens had much more pronounced fiber/matrix projection zones,, This can be attributed to theInterfiber distances being large enough to allow thematrix to achieve sufficient ductility to maximizefracture energy, through additional shear and subse-quent multiple fiber fractures Comparison of figs.10-b and -c shows the Increases In shear deformationduring impact of 1100 Al matrix composites allowed byincreasing the boron fiber diameter from 5.6 to 8 mils.The use of a mil boron in corilpo , ' Ites w-ith othermatrices also Increased their Impact strengths overthose previously reported for 4 mll boron. From theseresults,4 It may be postulated that the use of evenlarger diameter boron fibers could further increase theimpact strength of com posites with 6061 and 5052 Almatrices,,

Ref. 24 reported results from Charily' Impact testson boron, carbon, or glass fiber composites with resinmatrices of various tou ghnesst Calculations were made

tlORIGINAL PIE IS 13AV

-QF POOR QUALM

}r

a

to determine the relative contribution of fiber pull-out, shear delamination, and fiber fracture energies.Two-thirds of the calculated energy came from theenergy absorbed by fiber fracture, which was in turnproportional to the area under the stress-strain curveof the fiber. Glass fibers, with much higher strengthand failure strain, had the largest area under thestress-strain curve and gave the highest Charpy impactstrengthso Boron fibers were next, and carbon fibers,with the lowest strain and area under the curve had thelowest impact results. Furthermore impact strength wasIndependent of the toughness properties of the matricesdue to the overpowering influence of the fibers.

These results are significant because they showf

that in a brittle matrix system, the major contributionto energy absorption comes from fiber fracturingbComposite impact properties are an interaction of theenergy contributions of each constituent: the matrix,the fiber, and the Interface. However the strain andImpact behavior of each component are interrelated andmust be such that the full contribution from each canbe attained, Brittle resin matrices do not contributemuch to the energy absorbing capability of a composite,A ductile matrix, such as 1100 A1 0 can make a signifi-cant contribution to the overall impact energy by 3

absorbing additional energy by matrix shear as well asby allowing multiple fiber fracture. Thus it isvitally important to have a matrix with sufficientductility to allow the fibers to attain a greaterportion of their full strength and strain capability,

E t eCt 2f egl v Due to the an i sotrop i c nature ofcomposites, the transverse properties of unidirecti-onalcomposites may not be -high enough to withstand stressesencountered during_ component service. Angleply layupscan be used to improve the transverse properties;thistransverse improvement, however, is attained with aconsiderable ,penalty in longitudinal properties.

_Angleplies of ±7, ±15 +22o and ±30 degrees for 8mil B/1100 Al composites were studied. in addition,results from tensile and full-size Chirpy impact testsfor three angleplies with 1100 and 5052 Al matrix com-posites were reported in ref. 5. The first angleplywas (±45/0 ) , consisting of 50 unidirectional pliesin the central core with 25% alternating t45 degreeshells on eachside of the core. The second was(0/±22)'nT, and consisted of repetitive 0,+22,-22 degreeplies. The third was alternating ±15 degree plies.

ORIGINAL PAGE IS

OF POOR QUAI.i* . 14

a

w4

t

The longitudinal tensile strength of R/1100 Aldecreased with 'increasing angleply,, This reduction wascaused by a decrease in the elastic strain range and anincrease in nonlinear strain shown in the stress-straincurves of fig,, 4. The transverse modulus and strengthIncreased slightly with increasing angleply. SirnI I a rresults for B/1100 Al and x/5052 Al composites werereported in ref. 5, but with higher longitudinal andtransverse strengths due to different bondingconditions

Fig, 13 compares the area-compensated longitudinalimpact strengths of angleplied B/A1 composites from theiIASA-Le RC in-house program and from ref q 5 Full-sizeCharpy specimens showed a linear loss in impactatrength with an g lepl y , Because of the difference inbonding conditionsp the two sets of data are displacedfrom,# but parallel to, each other, Thus, the trendsfrom the two can be compared,, Unidirectional specimenshad higher impact strengths than an y of the angleplies,The reduction was iilinor up to t15 degrees,, The 7degree angleply had a minor loss in impact and tensile

ii strength compared with the unidirectional specimensIncreasingwhere tracture l occurred. Attlanglepl^ies t greater ^than

j

+15 degrees,the non-linear stress-strain behavior andlow strength allowed the composites to deform withoutapplying sufficient stress on the fibers to attain high-impact. The angleply specimens that did not bread., ±22 1and +50 degree, underwent considerable stretching and 1distortion during impact testing and showed alargeamount of shear, but the stresses required for deforma-tion were low due to the low flow stress of these com -posi tes. The filers were not strained enough to maketheir maximum contribution to the properties of thecompos ite , The maximum angleply that al lowe d the fiberproperties to be utilized was ±15 degrees,, In thiscase, the fibers fractured after* attaining sufficient_,.strain to give high stresses and impact energies,,

The ±45 shell-0 degree core configuration had thebest transverse strength and impact properties, butalso had the lowest LT properties ( ref. 5). The ±22, 0anglepl y gave slightl y lower TT impact and tensilestrengths than the ±15 degree. angleply, This was dueto the 0 degree fibers _which gave adverse results inthe transverse dire-ction. The best combination oflongitldinal and transverse impact and tensile resultswas obtained with the 15 degree angleply.

15

r

6^ y

a

x

MaJ!:ix p 11i1r3=eme" Another method of

improvin g; transverse strength is matrix enhancement,where a third material, either in foil or fiber form,is placed between the aluminum matrix foils to modifythe matrix properties.

Refs. Gil reported that the addition of 6 /ostainless steel wire,. oriented in the transversedirection, Increased TT Charpy impact strength of 4,2mil Borsic/6001 Ai composites from 1,5 to G40 joules

' (1.1 to 405 ft-lbs)¢ i: urther work (refs 13) showedthat LT Charpy impact strength of 4 mil R/G_O61 Al wasincreased by 60% to 26 joules (19 ft-lbs), with anaccompanyin g increase in transverse tensile strength,by using a dual alloy matrix of G061/1100 A]. it wassuggested that LT and TT impact strength and transversetensile streng th could be increased further by usinstitanium foil as matrix enhancement.

I, Results were reported In ref. 25 for diffusionbonded and adhesively bonded 5,6 mil a/6001 Al compos-ites, as well as 5.6 mil B/6OG1 Al hybrid compositeswith adhesively bonded 0.038 mm (040015 in,) thickTi-GA1-°4V foils, The area-compensated thin-sheet IzodImpact strength of adhesively bonded B/Al was increasedfrom 32 to 43 joules/sq m (15 to 21 ft-lbs/sq in,) forthe B/AI+Ti hybrid, Without hybridization, values fordiffusion bonded 4 mil B/6061 Al were 45 joules/sq m(22 ft-lbs/sq in.) and 49 joules /sq m (24 ft-1 bs /sqin,) for diffusion bonded 506 mil B/6061- Al. Furtherfiybridizati'on by adding graphite fiber/epoxy plies to6/AI+Ti hybrids increased area-compensated Izod valuesto 117 joules/sq m (56 ft-lbs/sq in4)o (These valuesshoul d be compared to the thin-sheet Izod resultsreported in this paper: 8 mil B/1100 Al-. 192 joules/sqM; 5.6 mil B/1100 Al a 89 joules /sq m; and 4 mil B/1100A]: 75 joules /sq m)o

Ref. 5 reported the use of Ti-GA1 m 4V foils with5,6 and 8 mil B/110-0 Al to determine the effect of ma-trix enhancement. Results showed that matrix enhance-ment- reduced longitudinal tensile strength 15% andreduced full-size LT Charpy impact stren g th by over 50for both compositeso The transverse tensile strengthwas increased from 65 dPa (10 ks i ) to 266 I4Pa (39 ks i )however TT Charpy impact strength was only increasedfrom 1.4 joules U ft-1b) to 441 joules (3 ft-lbs)oThis slight increase in TT impact_ strength did notjustify the sacrifice in ` LT impact, The same trendsheld for angleply B/5052 Al composites, The LT impactstrength was reduced by more than 50 while the TT

WE16

a {

.moo►..:

}

impact strength was virtually unchanged by titaniumfoil enhancementQ

The data obtained In ref,, 5 seem to differ fromother reported resultso it is generall y thought thattitaniumfoil Interleaves should Increase Impactstrength of B/Al. Titanium Is very impact resistant inthe unnotched condition, however a notch reduces thefull-size Charpy Impact strength from 318 joules (220ft-lbs) to 23 joules (15 ft-lbs). Thus in a notchedImpact test® titanium foil matrix enhancement shouldonly Improve Impact strength of composites havingImpact -strengths below that of notched titanium (2. 3joules). Titanium foil restrains the matrix fromshearing, thus making fracture and crack Initiationmore difficult, thereby Increasing impact strength ofbrittle compositeso It also provides delaminationplanes for low Impact compositese which rely ondelamination surface energy dissipation to ImproveImpact behavior, The matrix ductility restraintimposed by matrix enhancement foilse however,, willembrittle more ductile composites, such as 13/1100 Al',By not allowing the matrix to shear, this restraintwill not permit the fibers to attain their fullstrength contribution&

Corilparlson of the SEM fractograph presented infi

g. 14 with that In fiaz 9-b shows that the fiber/

matrix projection zones are broken up by the titaniumfoi I s,, The fracture is planar with much bare-fiberpull-out and no evidence of matrix shear ductility.

^ ^ 9—f- L t--i--ona . _tva - An unexpected direct on-al I ty ef fect reported 1 n ref,, 5 was the reduced i mpactstrength observed in full-size Charpy tests in theLT(s) direction,, The impact strength for the LT(s)geometry dropped as much as 30-50 1"o below the LTstrength,.

in diffusion bonding, matrix foils are placedbetween fiber layers and consolidated, Upon impacttesting of LT specimens, the crack must propagatesequentially through fully dense aluminum foils withweaker Al/Al interfaces separating the individualfoils. In the other case, LT(s), the crack, mustpropagate simultaneously across the entire number ofpl ies act! ng as a un

I t,,

If bonding were not perfect, the strength of thefoils, in the fully dense direction in the plane of thefoil, would probably be greater than the strength in

ORIGINAL PAGE IS 17OF POOR QUA=

l3 t

1

}

31i

t

^

it

the direction where the foils were bonded to eachother. This can be seen from the notch-initiateddelamination present in ful1-°size LT Charpy specimens(fig. 15-°a)Q After delamination, the specimen Mends byshear and acts in a ductile manner, resulting in highImpact energies. The opposite case,, LT(s), does notundergo this type of delamination below the notch (fig.15-b). SEM fractographs of high-energ y LT specimens,figs 9-be showed massive fiber/matrix projection zones.The LT(s) specimens (fl ab 16) showed less fiber/matrixprojection zone formation, The fibers are alined inIntact vertical planes and appear to show evidence ofbare-finer pull-out. The vertical planes are from theIndividual ply layup during consolidation The crackpropagation direction is normal to the edge of the plyand the fracture crack proceeds throughout all theplies simultaneously. Instead of having uniform pliesto deform sequentially by shear, LT(s) specimens mustfracture simultaneously through all the plies. Sincenone of these plies are oriented preferentiall y forshear, the matrix cannot shear and the fibers are notpermitted to exhibit their maximum strain capability.Thus, the impact strength of LT(s) specimens is reducedto that approaching a restrained, non-ductile matrix.

App lication of Improved Impact Technologyto Aircraft Gas Turbine Engine Fan Blades

The very large increase in pendulum impactstrength of improved i3/Al composites described in thispaper is very encouraging. The advanta ges of theimproved B/Al composites are shown In fig, 17, whichcompares current values with impact strengths ofprevious B/Al and notched titanium. These resultsprovide a basis for expecting that a significantimprovement in fan blade performance might be obtained.However the results of low-velocity pendulum impacttests on laborator y ,specimens do not necessarily meansatisfactory foreign object damage resistance torcomplicated fan blade geometries at high velocity fanblade operating conditions.

Blade-like shapes were fabricated, tested andreported In ref. 5. These blade-like specimens had aflat, untwisted alrfoil-like section and a splayed3-wedge root. The root was placed in a clamp and thespecimens were subjected to low-velocity impact tests.Specimens of ±15 degree angrepl y g mil B/1100 Al failedat the root-airfoil fillet after 'considerable shear,thus indicating that the matrix shear displacement tookplace in a manner similar to that observed in Chirpy/

18

B

}

1

1 l►

lzod thin-sheet and standard specimens, A limitednumber of high-velocity tests with blade-like specimenswere also performed. Fi g , 18 shows a D/Al specimenafter high-velocity ballistic impact with RTV siliconerubber simulated birds (ref,, 5)Q Specimens were able towithstand impact energies up to 250 joules (184ft-lbs), the [Maximum energy tested,, Specimens deformedby shear, with deformation primarily in the root area.No delaraination was observed and leached out fibersindicated no evidence of fiber breakage at the root.

Both low-velocity pendulum and high-velocityballistic Impact results are encouraging, Additionaltests are required, including single blade static FODtests, whirling arm 'tests, and full stage engine groundtests, Flight experience must then be accumulated todevelop confidence that ii/Al is ready for broadapplication to fan blades,,

A start has been made with this effort and theresults obtained thus far are very encouraging. Thesepromisinu results should serve to further thecontinuation of the development of B/Al composites toobtain the large payoff in performance gain, fueleconomy, and cost and weight reduction that compositematerials can provide when applied to fan blades foraircraft gas turbine enginesp

SUMMARY OF RESULTS AND CONCLUSIONSONS

The following results and conclusions wereobtained from studies to improve the impact propertiesof diffusion bonded B/Al composites t.

I. Pendulum impact test results of Improved B/Al were`higher than notched ti tanium and appear to be highenough to give sufficient foreign object damageprotection for consideration of B/Al for applicationto fan and compressor blades in aircraft turbojetengines.

2, impact strength of B/A1 can be improved by properchoice of fabrication temperatures, Processing atbelow optimurii temperatures causes Impact strength tobe reduced by B/Al or Al /Al interface delamination4Above the optimumf impact strength would be reducedby excessive reaction at the fiber/matrix Interface.In this case the bond strengths are in excess ofthose required for best impact resistance,

I,

^y

i

4x

X ^ x

3 impact stren gths of composites with an 1.100 Almatrix. are si gnificantl y higher than with 2024, 5052and G p G1 Al matricesi More ductile matrices allowadditiona l energy absorption throu gh sheardeformation and multiple fiber breakage,

4e Larger diameter boron fibers increased impactstrength. They provide larger interfiber spacing,allowing the matrix to act in a more ducti le mannerand permit the fibers to attain a greater portion oftheir full strength and strain capability.

S. The LT(s) impact strength (notched side parallel topressing direction) was lower than the LT impactstrength (notched side normal to pressingdirection),

G, Transverse tensile and impact properties can beIncreased through the use of anglepiy fibers. Theoptimum angleply for impact resistance appeared tobe about +15 degrees:

k 7. Matrix enhancement, using titanium foil interleaves,reduces the longitudinal impact strength of ductile,high impact strength a/Al composites,,

8. Thin-sheet I zod and Charpy impact tests can be usedfor ranking purposes to compare impact propertieswith full°-size Charpy tests, but the quantitativeresultsof one hype of test cannot be extrapolatedto another*

REFEREiiCES

` 1^, Norhut, T. Ja "Fatigue Tolerance of Damaged MetalComposite al ad i n g, "

Jmg Qgf omUosite. Alat€r i a l s5211 Ate. l € Vehicles &a4 Rronul s i on 5ystgULA

Advisory Group for AerospaceResearch, London, 1973, pp , 26.

2.

?eJong, A, Pa: "Their Airs pace or Ours?" Shelljw i Qj_1 M ACXMO Noa 290, 1970, pp, 2-7.

30 McDanelso D, L, and Signorelli, R^ Au. "'Effect ofFiber' Diameter and Matrix Alloys on Impact-Resistant Boron/Alum inum Composites," INASA TND-8204, 1976,'

4, McDanels, D, L. and 5ignorel l i R,, A # "Effect ofAnglepl y and Matrix Enhancement on Impact-Resistant Boron/Aluminum Composites'," NASA TN

20

D-8205,, 1976.

5 Q Mel nyk, P. and Toth, 14 J.: "Development of ImpactResistant Boron/Al uminum Composites for TurbojetEngine Fan Blades," NASA CR-134770, May 1975,

G, Kreider, K, G,, Dardi, La and Prewo, Ka: "Metal14at r i x Composite Technology," AFML-TR-71- 204,December 1971

7, Dardi, L. E. and Kreider, Ka G,,.- "The NotchedImpact and Flexural Behavior of Boron-Aluminurti,"Fa Il u ^, L4 2"I a The Meta 1 1 u rb i ca 1Society of RIME New York, 1973, pp. 231.270,

8. Hoover, W. R. and Al I red,R,,, E, *. "The Toughness ofBorsic-Al. Composites with Leak Fiber-metal Bonds,"Report SC-DC-7144G7, February 1972, SandiaLaboratories, Albuquerque, N, 11,

9 0 iiancock, Jo Rn and Swanson, Ga D, "-Toughness ofFilamentary Boron/Aluminum Composites,"' CompositeMaterials: aILLE1& s" Deli. n .. STL-moo AmericanSociety for Testing; and Materials, Philadelphia,1972, pp, 299-310a

10. Coope-r, G, A. and Kelly, A,,: "'Tensile Properties ofFibre-Reinforced Metals: Fracture mechanics,""fnecho Phvs, g1 ids, Vol , 15, 1967, pp, 279-297.

11. Wilshaw, T4 R, and Pratt, P. La: "On the 'PlasticDeformation of Charpy Specimens Prior to GeneralYield," Phvsa 11, Vol a 14 1966, pp,7 - 19a

`• 12 0 Philli p s,, Da C, "The Fracture Energ y of Carbon-Fibre Reinforced Glass,"r A. JA e rials .Sci # Vol.7, 1972s pp . 1175 -1191 a

13. Chaiii1s, C. Ca, Hanson, M,, P. and SerafIni T. T."De signing for Impact Resistance with Unidirection-al Fiber Composites," 14AGA TN D-6463, Aug , 1971,

14. P i ggott, Ma R. : "The Effect of Aspect Rat i o onToughness In Composites," bYA. Solo .ids,Vol. 9, 1974, pp. 494-502a

15. Kelly, A p ; "Interface Effects and the Work ofFracture of a F ibrous Composite," Proc,, .la. ;,,tom,

i

YLondon), Vol, A-319 01970, pp: 95-116,

f' 21

f .... _... .. ^..... ....yy

k ^

! x jY i1]

gi

f

pg

i °^

16, ilancock # J. Ri and Shaw, Gn G.: "Effect of Filariient-Matrix interdiffusion or) the Fatigue Resist-ance o f Boron-A1umi num Composites," Composite lit-erials. _T&stinr g" Dgsi .pn . aT - 4 , AmericanSociety for Testing and Materials, Philadelphia,1974 0 pp. 497-506.

17 Prewo, K.: "The Charpy Impact Energy of BoronAluminum,." jm gamoos i to lateiK^.l.^. Vol, G, Oct.1972, pp^ 442-455

18. hoover, 11. R. and Allred., Rk Eo ; "The DynamoFracture Behavior of Borsic-Aluminum Composites,"Report SC-DC-721080, June 1972, SandiaLaboratories, Al buquerque, N. M.

19, Coope r, G, A. and Kelly, Aa; "Rol e of the InterfaceIn the Fracture of Fiber-Composite I+later i a l s,"interfaces in Comnosite2 ,, ,_,TP:_45.,, American Societyfor Testing and Materials, Philadelphia, 19G9, pp,90-10G

20, Coope r, G. A^,; "The Fracture Toughness o f Compos-Ites Reinforced with Weakened Fibres," SL.,Mgt,eri al sScience, Vol. 5, Aug. 1970, pp. 645-b54.

21. Harris, ; "The Strength of Fibre Composites,"Cor,nos rtes, Vol. 3, July 1972, pp 4 152-167,

22, Helfet, J,, L, and Harris, B. "Fracture Toughnessof Composites Reinforced with DiscontinuousFi-bres," J. Materials science s Vol ^ 7, May 1972,pp . 494-4980

23o Jech, R,, W. and Signorelli, R,, As p "The Effect ofInterfiber Distance and Temperature on the CriticalAspect Ratio in Composites," NASA TN D-4548, May19680

24. Novak, R. C. and DeCrescente, Mo A.: "impact Re-€ havior of Unidirectiona l Resin Matr ix Composites

Tested in the Fiber Direction," Compositetlaterials: Testing aild.Dgliza, ,SST -497, American

Society for Testing and Materials, Philadelphia,1972, pp , 311-323-0

25. Chami s, C, C., Lark, R. F. and Sul 1 ivan, TQ L."Boron-Aluminum - Graphite/Res-in Advanced FiberComposite Hybrids," NASA TN D -7879, Feb. 1975,

^:, 22

WUt ^ 2

2(7

^ a=o X ^ 2

o

uoc ^ X 333 ^W

^ O O OC7

J

W^ W

< m

VW y,

ag

W ` r N W

C x 2> 1< z < ^ OC N

udca^

N

^o0.

tCD

0

u

Sd3

LD

i

PI{17C1.,'DING PAGE BLANK NOT FII.M

ORIG&QFPpo QU GE IS

^Ll7i

9

^i E

^rG1

2 O U

VlW

CL C:)

I

Nt

Paz^;^^^.^^ ^

R QT

iQT

r^^ g^ ^ g

>o

^ r

^ c

^s Y

ISM 'N1')N]hlS DISN11 ]1VWllln 5 IS

^.1 1 1 ^ 1 1 ^ ^ ^

edW 'HONAI S 311SN11 31 V6C Ifl

W`V

M^W

ri

201)r-

160

eoo ; 1 ^'0s NN

200

0

12w

low

. rd ANC1mv

StRAIN, KRCINT

Iigure 4 Stress drain curves d 6 mil {,1100 Al composites, Ilsledin the longitudinal direction.

f

10015

5'152 A 100

aN 10

N so NN NVAI

IIUAI } 220iGlfplr

0 1 1 1

0 .5 1 1.5STRAIN, PERCENT

Figure 5. Stress strain curves of 8 mil B Al compos-ites, tested in the transverse direction.

600175 — TYPE OF FAILURE

1100 80 –NO O FIBROUS

i • DELAMINATION

0z 500

°C = 150 ^-- O1000 °, N 10O

FIBROUS_t g N \ i All ORES

800 g N u

<125

CL

^\

•Nu 300 -FULL-SIZE = 50

a b00 cL CHARPY i c100 O

o 40 H• ^DEIAMINATION

<^ 200L) z FAILURES

400 z < li' 15 • J!^ ^' or •

^S THI% S40a 30.. I-j

< 200 < 100r CHAR PY

of

5U•

l I 1 _J

or 100 125 750 715 800aBONDING TEMPERATURE, K

^THiN-SHEET i10D

o U_ 1 i J l 1 1 t4.0 5.6 8.0 800 850 900 950

FIBER DIAMETER, MILS BONDING TEMPERATURE, OF

Figure 6. Area-compensated longitudinal impact strengths Figure 1. Effect of bonding temperature on areaof unidirectional 1100 Al composites from different types compensated thin sheet longitudinal a0d impactof Impact tests, strength of 8 mil 811100 Al composites.

ORIGINAL PAGE 5OF POOR QUALITY

f"

I

I

DUCTIL

ut

E MATRIX

I I^

(b)

BRI"L( MATRIX

MATRIX

f igure S. Crack propagation in composites with matrices of

ditto rent ductilifies tr el. 191.

00r

tit S.n V It 81505? Al.

Aj4

'./st t^

*v

L

IN 5.o M11 Bill(Xi Al. 40 8 Mll. B/IIUU Al.

Figure 9. - Failed lull-site LT Chirpy impact test specimens showing

effect Of Various matrices and fiber diameters on protection zones

in B/Al composites, lRel. 5,1

(al 5.6 MIL B15052 Al

ib) 5.6 MIL B1100 At18 J 113 FT-LB). 64 J 147 P-LB).

v(c) 8 IVIL B/1100 At

-473

96 J 171 FT-LB).

i 1 ,jure 10. - failed full-size LT Charpy impact test specimens. iRef. 5. 1

ORIGINAL, PAGE ISOF P001i QUAj,rTy

70 -10D

SPECIMEN END

60— SHEAR c., 15 DISPLACEMENT o

50 > oG 1 . 1100 A1,0°

40 — S0^ ° )1 6061 A1,0°O / 5052 A I, 00O IlSIC-TI16'41,0°

i nI

op /1SIC Tilb U, t Is°G1 -1100 A I, I l5°

' [2 8-1100 At. Of ?20010 IQ 1-1100 At, ot450

o t — 111 jo .127 2sA .381 508

SHEAR DISPLACEMENT, Chi

0 .050 .100 .150 .200SHEAR DISPLACEMENT, IN.

Figure 11. Shear displacement at the ends of various failedfull-site t T Charpy impact test specimens (rel. 51.

I .

A

Figure 11. - Fracture region of failed uni-directional 8 mil 811100 At thin-sheetItod impact test specimen showing multi-ple finer hreakarte,

bw

1200

r

o ^No 1000 oa

TEST TYPE NASA REF. S

FULL-SIZE CHARPY O OTHIN-S ►EET CHARPY O -T HIN-SHEET170D L -

p ,IGIN,NL PAGE LSOF POOR QUALM

Y ►

800

u 6w 300

a a

o c

4w :'00z _

O L O }^

' V^ ------±DID NOTW 200 100

FAILi

oJ

0 t t15 t22 t30ANGLEPLY. DEG

Figure 13 Area-compensated lonq;tudinal impact strengthsof 8 mil 8 1100 Al angleply composites

ter.` ,'^^^'^"^' ~+'fit ^'^r • ¢^

,KW• ..

Aq\ is _ i • ^

.iv-6

Figure 14. SEM fracivyipt . r,il 811100At composite with titanium Iwo enhancement. LT impactstrength, 41.5 J 130 WIN. tRet

I

(a) li; IhfPACf Si RIM1I,Iri, '.ACT SIRF.NGTH, 41.5 J96.5 J Ill FT -LO).

Figure 15. Failed full-size LT and LT(s) Charpy impact test specimens.(Ref. ].)

?RI N 1'A

QUA Q11GE

t,yure 16. - Fracture surlac.lLTtsl test specimen. Impact sttRef, 5.!

S •- Mi

G

iil0 Al fu!I-sizeJ 135 ft-lb).

6

75 — 100

60 ID

o

u45 G 60

<

d 30 4D

uV° 15

p20

UNI TITANIUM IMPROVED,., DIRECTIONAI ALLOY UNI

BIAI DIRECTIONALBIAI

Figure 17. • Frogress in Impact improvement.

O R p ^^ pQ0?

QU^A^GF^LS,"1Y

Figure 18. - B!AI blade-like shi, l* specimen attar high-velocity ballisticimpact testing. IRef. 5.1

NASA-Lewis