We are Nitinol.™
TiNi Shape Memory Alloys
Duerig, Pelton
Materials Properties Handbook Titanium Alloys pp. 1035‐1048
1994
www.nitinol.com 47533 Westinghouse Drive Fremont, California 94539 t 510.683.2000 f 510.683.2001
25 March 2009 Letterhead (scale 80%) Option #1 NDC Business System R2 Ti-Ni Shape Memory Alloys 11 035
I Ti-Ni Shape Memory Alloys T.w. Duerig and A.R. Pelton, Nitino/ Development CofpoIation Effect of phase transformation This datasheet describes some of the key prop erties of equiatomic and near-equiatomic tita· nium-nickel alloys with compositions yielding ~ C~ i shape memory and superelastic properties. Shape memory and superelasticity per se will not be reo D viewed; readers are referred to Ref 1 to 3 for basic Manensile Auslenile {pa.ent) information on these subjects. These alloys are Typically 20 ~ commonly referred to as nickel-titanium, tita· D- M, .. Manensl1e stan nium-nickel, Tee-nee, Memorite.... , Nitinol, Tinel"', temperature and Flexon"" . These terms do not refer to single al· Heating ' j A, lIAr ~ Manensite linish loys or alloy compositions, but to a family ofalloys .. temperalure with properties that greatly depend on exact com A" "Stan of reverse positional make-up, processing history, and small " transformation of manGnsHe A. = Fin iS!' ot lI!Verse ternary additions. Each manufacturer has its own transformatIon 01 martens;t series of alloy designations and specifications within the "Ti·N i~ range. A second complication that readers must ac Temperature ->
knowledge is that all properties change signifi Sd"Iemane "U$tra~OI'1 of It.. ",Heels on a phase transformation on cantly at the transformation temperatures M Mr, !tie phySical prope~s ofT,·Ni. All ~t properties eilibit a dos· ~, and Ar (see figure on the right and the section" continully, d'\ar.JCt~rized by !he translormation tempe Product Forms Titaniwn-nickel is most commonly used in the (1020 oF). Bnd form of cold drawn wire (down to 0.02 mm) or as Applications for t itanium-nickel alloys can be Applications barstock. Other commercially available forms not convenient ly divided into four categories (Ref8): yet sold as standard product would include tubing Free recofJery(motion) applications are thoS(' i:l (dov,n to 0.3 mm 00), strip (down to 0.04 mm in which a shape memory romponent is allowed thickness). and sheet (widths to 500 mm and thick· to freely recover its original shape during he:!! nesses down to 0.5 mm). Castings (Ref 4), forgings ing, thus generating a recovery strain (Rcf9 . and powder metallw-gy (Ref 5) products have not Constrained recovery (force) applications ar!! yet been brought from the research laboratory. those in which the recovery is prevented, con· Typical Conditions, Titanium·nickel is most straining the materiru in its martensitic, or commonly used in a cold worked and partially an· cold, form whil~ recovering (Ref 9). Although nealed condition. This partial anneal does not re no strain is rer,) vered, large recovery stress,'~ crystallize the material, but does bring about the are developed. These applications include fa s· onsel of recovery processes. The extent of the post teners and pipe couplings and are the oldest cold worked recovery depends on many aspects of I1J1d most widespread type uf practicru usc. the application, such as the desired stiffness, fa· tigue life, ductility, recovery stress, etc. Fully an Actuators (work) applications are those in nealed conditions are used almost exclusively which there is both a recovered strain ar.d when a maximum Ms is needed. Although the cold stress during heating, such as in the case of n worked condition does not transfonn and does not titanium-nickel spring being warmed to lift n exhibit shape memory, it is highly elastic and has ball (Ref10l.ln these cases, work is being done. been considered for many applications (Ref6). Such applications are onen further catego' R esponse to Heat Treatment. Re<:overy ri zed according to their actuation mode. e.g .. processes begin at temperatures as low as 275 "C electrical or thermal. (525 oF). Recrystallization begins between 500 and 800 °C (930 and 1470 oF), depending on alloy com Supereiasticity (eTU!rgy storage) refers to the position and the degree of cold work. highly exaggerated elasticity, or springbacK. Aging of Wlstable (nickel-rich) compositions observed in mRny Ti-Ni alloys deformed abo..-e begins ot 250 "C (525 "F), causing the precipitation A. and below Md (Ref 11). The fundion of t I:.e of a complex sequence of nickel-rich precipitates material in such cnses is to store mcchan: ::u (Ref7), as these products leach nickel from the ma energy. Although limited to a rather small i.e trix, their general effect is to increase the Ms tem perature range, these alloys cun delivcr 0' 'r perature. The solvus tem;leratw-e is about 550 °C 15 times the elastic motion of a spring steeL Special Many shape memory-related properties are temperatures. superelasticity, etc.). Sume proper Properties discussed in subsequent sections (transformation ties, however, are ~ tric tly peculiar to shupe men!- 1 C361 Advanced Materials ory aUoys and cannot be conveniently categorized Free recove ry behavior in standard outline fOrIDs. The more important of these properties are discussed below. Free-recoverable stram in polycrystalline • titanium-nickel can reach 8%, but is limited to a maximum of6% ifcomplete recovery is expected. • Applied stresses opposing recovery reduce recoverable strain. Clearly, stronger alloys will be affected less by opposing stresses. Work output is mrucimized at intennediate stresses and strains. Recoverable stresses generally reach 80 to 90% of yield stress. In fact, alloy behavior depends on numerous factors, including the compliance of the resisting force and the constraining strain (Ref 9 and 12). 1)rpical values are as follows: 5 10 lS Condition Recovery 5tree~. MPa TOlal deformalion strain, 'Y. " AMQlcd b;.ntoo;k 'T1·NI-Fe be rslock wi\tl SO a1.% NI8/1d 30/. Fe fuU)' annealed. tested Co~~~b~k~cd "'" OISOO ·C(930"F) '''' In unlilldal tltf"lslon. An", deforming 'T1-Ni to various lotal slrnil'lS Ix Cold ""Ot\(od wi~:umr:.oJod ~l '000 axis). the malerial S!>f\ngs bae!< to \tie ptastic strain I~"s shown by 4OQ °cnso·F) \lie open cirdes. A~er hea~ng above A,. most oI"e strain Is recov· ered, but $OffiO amonla pertlSIS. lhedifferenoe between the plas· tie slrain IW'd the amnesia is the reoove Effecls of opposing stresses on recovery strain Work output of a li·NI alloy Applied stress. ksi Applied stress. ksi o 10 ~ ~ ~ M W ro ~ o 10 ~ ~ ~ ~ ~ ro ~ O' ~~~~ __~ __~ __~ --J o t OO 200 300 400 500 600 Applied stress, MPa A~plied stress. MPa T\.Nl-Fe tlarstock with SO at.% Ni and 3% Fo fUly anr.&aled, tested To·Ni·Fe barslock wil:l150 al.~~ Ni and 3"10 Fe in II wor!<·harclened in unia;cialleflSiot>. condition, tesled in unill.lllatlens4on. Source: J.L. Proll and T.W. Ouertg, Fngneering AspecI$ 01 Shap6 Source: J.L Prof! and T.W. Outlig. Enginee Chemistry and Density Density, 6.4510 6.5 glcm3 Titanium-nickel is extremely sensitive to the memory alloys in the range of19. 7 to 50.7 at. %. Bi precise titanium/nickel ratio (see figure below). nary alloys with less t han 49.4 at.% titanium are ~neral1y, alloys with 49.0 to 50.7 at.% titanium generally Wlstable. Ductility drops ropidly A8 are commercially common, with superelastic al· nickel is increased. loys in the range of 49.0 to 49.4 at.% and shape Binary alloys are commonly available with :\Its Ti·Ni Shape Memory A lloys 11 037 temperat.ures bet.ween -50~ and +100 ·C (-58 to Effect 01 composition on M. 212 -F). Commercially available temary alloys are available wit.h M.. temperatures down to-200·C ~ (-330 -F). Tit.aniwn·nickel is also quite sensitive to alloying addibons. ,. "'" Oxygen forms a Ti.Ni20.., inclusion (Ref 13). , ~ •, ,.. tending to deplete the matrix in titanium. lower p , M , rct.o.rd grain growth, and increase strength. ,; • • ", • s ~ Levels usually are cootrolled to <500 ppm. Nitro ~ • .P ", '00 gen forms the same compound and has an additive & ~b' i ~ effect to ox.ygen. ~ .'" • Fe. AI, Cr, Co, an.d V tend to substitute for • .5( • • nickel, but sharply depress Ms (Ref 14 to 16), with ·'00 V and Co being the weakest suppressants and Cr .'0< .' the strongest. These elements are added to sup , ·200 press M, while maintaining stability ar.d ductility. . '5( Their practicu.l effect is to stiffen a supereJastic al· ., os ., 50 loy, to create a cryogenit: shape memory alloy. or to Nick,l. al.% " " " increase the separation of the R-phase from M. lemptralu,os 111 niek4ll ·til;nivm alloys alII extremery s.er1silive 10 martensite. CIOI'IIposIllonl! Y8~a~on, per1ievlarty al rll!1'er nIc~el conlenls_ Pt aDd Pd tend to det:rease M, in small quan· SoYroo: K.N. Mollon, E~ Aspects Of SI!SOe Mer7IOt'Y Al tities (-5 to 10%), then tend to increase :'vI •• eventu loys, ToW. Ouerlg (II iii., Ed., avn(ll'WO<1h-~. 1990. p 10 ally achieving temperatures as high as 350 °C(660 ' F)(Refl7). Zr acd Hi occasionally have been reported to martensitie strength. Nb is added to increase hys.· increase M., but are generally neutral when sub· teresis (desirable for coupling and fastener appli· stituted for titanium on an atomic basis. cations), and copper (Ref 19) is added to reduce Nb Illld Cu are used to control hysteresis and hysteresis (for actuator applications). Central Portion W(,lgh : Percent Nickel ofTi-NiPhase UOO ..,, .. Diagram J AS)(. IS87 0 $4"" L i 1 300~ o 148., , o I'P Phases and Structures •. Crystal The high·temperature austenitic phase (Jl>' has a is worth noting that there is aho a "trnnsitioll~ Structure B2. oresCI ordered structure with no ",,3.015 A. The structure that preceded the mllltcnsite. called the most common martensitic structure !BIg') hu a R.phssc with n rhomboheUral StruL1.ure (Ref 2U. complex monoclinic structure with 4 = 2.889 A. 0 = Although this R phose e:<.hiiJils a number of in U'" 4.120 A. c =4.622 A, and ~ = 96.8" (Ref 20), The Ms ('sting properties, it will not be reviewed e:dem· can range from <-200 to +lOO°C (-328 to 212 ~ F ). It sivcly here. 10~ I Advanced Materials Ti me-lemperature-transformation curve OOO--0~~ TiN;------~'100 • 0 • TiN; . Ti!!Ni!.,," r~iJ 0 TiNi • 11zNiJ .. rlN~ • • TiNI "" TinNi!. 0 TiN;. r~~ '''''' • • 0 0 ~ • TiNi . TiN~ 0 0 • • • • • . • . •• ~ . • HOO ~ • • ·0 • 0 • • • • i 0 • . • • , • t200 f r • • ''' .. • ~ IlOO • • • • • J • • '000 500 • • ..., 0.01 0.' 10 100 "00 "000 Aging dm" h r.,...t,""*,,tur&-transfonnalion CUMlIor Ti-5 1Ni, 'o'dIictt shows p-ecipiIallorl reacIiOo& IS a lunctiOn 01 ten..-arure and lime. Source: M. N"1Shida. C.M. Wa~ andT. Horma, MetaII. ThIns.. A. Vol 17 , 1986, P 1505 Transformation The T-T-T diagram shows the aging reactions time increases at a constant temperature. Thesc Products in tulstable (>50.6% Ni) titanium-nickel alloys precipitation reactions can be readily monitored (Ref 7). in general, TiNi -t TiuNil", -+ 112Ni3 ~ via transformation temperature or mecharucaJ TiNi3 as the aging temperature increases or aa property measuremenle. Physical Properties Damping Internal friction and damping of titaniwn cooling through the M •. These usuaJly high damp Characteristics nickel alloys are dramatico.lly affected by tem ing characteristics (Ref 22) havc been studied for perature changes (see figure on left). Cooling (or some time, but have not been used un a commer heating) produces peaks, which correspond to the cial basis due to their limited temperature range trnnsformation temperatures. At higher tempera and rapid fatigue degradation. tures, a very sharp increase is observed during Elastic Dyno.m.ically measured moduli (Ref23 and 24) right). Tytca1 values of clastic moduli are 40 GPa Con stants change markedly with the martensitic transfor (5.8 x 10 psi) for martensite and 75 CPa nO.8 x mation and premartensitic effects (see figure on 106 psil for uustenitc. From a pmctical point of TI -NI-Cu alloy damping characteristics DynamiC Young's modulus vs temperature Ttmptratur,. OF Temperatu .e. "F ·200 -100 0 100 200 300 400 ".~~-".~>oo~-". '~OO"-~'"--"' ;OOC-"2~00"--"~~--, 10" !"" .. • I:J ~ •0 • 12 ~ .;• -/---':f-----=---l .; ~ ~ , II i .~ , - 0 --- --'--- "1 >• ., L~ ____-"~ ______"- ______-',- ____-,J ·200 -100 a 100 200 _170 ·10 30 Temperature. "C Tempenlture. ·c '" A: n.s6Ni (wr"o). 8 : 44.7T.. 29..3N1 ·26Co (wI'I'.). C: 44_9T.. S1.7Ni- Internal friction of 44.7T.. 29.3N"26 Cu (wt'!\.) du~ng coofing wi", 3.4Fe ('<01%). rne6suretnet\t trequeney of - t Hz. SoJroe: O. MerOer. K.N. Melton, R. Gontlartl. i!f1d A. Kulik. Proc..InL Source: O. M,rtilll' arxl E. lOr6k, J. Phys., Vol C·4 (No. 43). 1982, p CMl. StJid.-Sotia PPIa:M TllWroommions. ....I . AAr01SOl. D_E. 1..aug> C~'" lin, A.F. s..6fka, andC.M. Wa.,man, Ed.AtME. 1982, P t259 T i~ Ni Shape Memory Alloys / 1039 view, however, modulus is oClittie value; apparent Electrical resistance vs temperature elasticity is more controlled by the transfonnation Temperature, OF and by mechanical twinning. Poisson's ratio, Jl., is -200 -I" 0 200 0.33 (Ref25). '" M, "I I Electrical General values fo r electrical resistivity of the A • , Resistivity two primary phases are as follows (Ref26): ~ p (martensite) = 76 x 10-6 n · cm • M,~ p (austenite) =82 x 10-6 n· cm •• T, ~ Variations in resistivity with temperature are • , complex functions of composition and ther • ~ ~ , • momechanical processing (see figure). Note also • r-J- M. T, T-, ~" 1 the pronounced effect of the R-phase on resistivity. "I 1 T- • • I ' c Magnetic Magnetic susceptibility also undergoes a dis· '\~ Characteristics continuity during phase transition (Ref 26). Typi. ·170 · 140 · 110 ,80 ·SO ·20 10 40 70 100 130 cal values are: Temp erature, · C X (manensite) = 2.4 x 10-6 emu/g X (austenite)::: 3.7 x 10-0 emuJg Electrical resistance \IS ternperalure <;IJI'Ves lor a TI-50.6N; (at."-) al· loy IhlIt WIIS thermomecha"ie8~y treated as ;r'!d;cated. A: Ouend1ed tram l ooo ·C (1830 .F). B: Oue;'1dledfrom l(xx)'C (1830 'F), aged at400 "C (750 ' F). C; Dlr&dIy aged at 4OO "C (750"F). T R;$ thetran· s1tlon tempertltUfe from austenite lQ the rOOmbohedral A pt1as.e. r R Corrosion is the shifted transition temperature IrQ(ll prcx:essing eMects. Arbi trary units lor ~eef1ical resistance. Soun::.: S. Moyauki 3nd K. Otsulo.a . Melal. Trans. A.VoI 17. 1986. Titanium·nickel generally fonns a passive ,53 TiD2 (rutile) surface layer (Ref27). Like titaniwn alloys, there is a transition temperature of about 500 °C (930 oF), above which the oxide layer will be dissolved and absorbed into the material. Unlike titaniwn alloys, however, eoa case is fonned. Tita nium-rockel will also react with nitrogen during heat tnatments, fonning a TiN layer. General The rest potential of titanium~nicke l in a dilute tacks titanium-nickel at 5.5 mpy. Corrosion sodium chloride solution is around 0.23 V (SCE), Feme chlorid e (FeCI3) at 70 "C (160 OF) and which compares with 0.38 V Cor type 304 stainless 8% concentration attacks titanium-nickel at 8.9 steeL This puts titanium·nickel on the noble or mpy. Titaniwn·nickel is attacked at a rate of 2.8 protected .side of strunless steel in t he galvanic se mpy in a solution ofl.5% FeCla with 2.5% HCI. ries. A passive oxide/nitride surface film LS the ba· Hydrochlo ric acid (HC!) has a variety of ef· sis of the corrosion resistance of titani1,;ltt·nickel fects on the corrosion of titanium· nickel alloys de all oys, similar to stainless steels. Specific environ pending on temperature, acid concentration, and mcnts can cause the passive film to break down, specific alloy composition. With 3% He l at 100 "C thus subjecting the base material to attack. A 8um ~ (212 OF) and a range of alloy compositions, the rate mary ortitaniwn·nickel reactions in various envi of attack was as low as 0.36 mpy and as high as 3.3 ronments rollo ws (Re[28). rnpy. At 25°C (77 OF) and 7M solution, titanium· Seawatcr. Titanium-nickel is not affected nicke l~iro n alloys can lose up to 457 mpy. when immersed in flowing seawater; however, in Nitric acid (HND.1) is mo re aggressive toward stagnant 5eawater, such as found in creviccs, the titanium-nickel than ty pe 304 stainless steel. At ~rotcctive film can break down, which results in 30 °C (86 OF), 10% liND3 attacks at a rate of2.5 x pitting co rTosion. 10-2 mpy; 60% solution attacks at 0.25 mpy; 5% Acetic acid (CH3CDDH) attacks titanium HND3 at its boiling point attacks at 2 mpy. nickel at a modest rate of 2.5 x 10-2 to 7.6 x 10-2 Biocompatibility studies have been oon mm/year (m py) over the temperature range 30 °C ducted in various media chosen to simulate the (86 oF) to the boiling point and over the concentra conditions of the mouth and the human body. In tion range 50 to 99.5%. general, no corrosion of titanium· nickel "!loy.'! has Methanol (CH3DHl appears to attack tita~ been reported. For example, in tests where cou' mum nickel only when diluted with low concentra· pon! oftitaniwn-nickel were sealed at 37 °C (97 "F) tions of water and halides. This impure methanol for 72 h, the mass corrosion rute was on the order solution leads to pitting and tunneling corrosion of 1~ mpy for such media as synthetic saliva, :lyn· simi lar to that found in titanium alloys. thetic sweat, 1% NaCl solution, 1% lactic acid,.llld Cupric chloride (CuCI2) at 70 "C (160 oF) at- 0.1% HNaS04 acid (Hef29); see also Ref30. Hydrogen The interaction between hydrogen and tita· in excess of 20 ppm by weight can ue considered Damage nium·nickel is sensitive to both concentration and detrimental to ductility, with levels in ~xce!\s o f200 lemperat~e (Ref 31). In general, hydrogen levels ppm severely impru!ing. Undcr certain conditions, 1040 I Advanced Materials hydrogen can be absorbed during pickling, plat absorbed in hydrogenated waleI' at elevated tem ing, and caustic cleaning. The exact conditions re peratures and pressures, such as would be found quired for hydrogen absorption are not well de in pressurized water reactor primary water sys· fined, 80 it is advisable to exercise care when terns. Relatively short exposure times have been perfonning any ofthese operations. shown to produce hydrogen levels well in excess of Substantial amounts of hydrogen also can be 1000 ppm (Ref32). Thermal Properties Heat Capacity A typical plot of specific heat (C ) versus tem SpecifiC heal (Cp) perature fo r a 50.2% Ti alloy (see figure) shows a discontinuity at the M! temperature of90 °C (195 Temperature. ' F OF) (Ref 3). The peak and onset temperatures for • 00 -200 0 200 400 600 900 3 , , , the peaks are often used to characterize the trans ,i ,: I I i" fonnntion temperatures of an alloy. Care must be 25 - -,-- ,-- t-- -t- _ L~~ _- 5 taken however, (Ref 33), because the presence of -- , , , , I an R-phsse prior thermal cycling, and residual , , , stresses from sample cutting can tend to compli , - - I I , " cate the CW""Ves and introduce spurious peaks. , I 5 , ! ! " Latent The latent heat of the martensitic transfonns ! , I I Heats tion strongly depends on the transformation tem " : I I " perature and stress rate (doldT) throug h the for : , , mula ,~ i doldT = iiliKOF.n I I I , Typical values for 6H are 4 to 12 caUg and val ·270 ·180 ·SlO 0 90 1110 210 360 "50 ~O ues for do/dT range from 3 to 10 MPaJOC. T,mper.uu-e. ~ 8S: The latent heat of fusion can be expressed SpedIc tltat (II T0-49.8Nj (at ~.l . with a sharp peak ~ he spociIic llH =-34 ,000 J/mol (Ref34). heiLt at 90 "C (195 ' F) GOTeS4)Ol'ld"ng tl t-.e M. ~ture _ Scurc.: C.M. hd< __• H.J. W;q>er. and R.J . wasilewski. NASA fIepott. NASA-SP5110. 1912 Thermal The thennal coefficient of linear expansion can Expansion be expressed 8s(Ref23): (l (martensite) = 6.6 x lO--orC a (austenite) = 11 x l~I"C The volume change on phase transformation (6V) (from austenite to martensite) is -0.16% (Ref 35). Transition Temperatures Melting Point Martensitic Characteristic transformation temperatures tion temperatures are as follows. Applied stre!lses Transformation depend strongly on composition (see table on increase transformation temperatures according Temp eratures next page and the previous section on chemi, to the stress rate (see the next section on tensile try). Typical hysteresis widths range from 10 °C properties). Martensitic deformations increase (18 OF ) for certain titanium-nickel-copper alloys, the stress-free A, temperatures, particularly in al to 40 to 60 °C (72 to 108 OF ) for binary alloys, to loys with low yield stre!lS(>S. The increase is tempo 100 °C (180 OF ) for titanium-nickel-niobium alloys. rary, returning to the previous value after the first Transformation temperatures are measured heating cycle. Increasing cold work tends to reducc by a number off.echniques, including electrical re transformation temperatures. The R-phase trans· sistivity, latent heat of transformation by differen formation temperature is much more constant tial scanning calorimetry, elastic modulus, yield thao those for martensite, typically 20 to 40 °C (68 strength, and strain. However, the most useful to 105 OF ) in binary alloys. measurement technique is to monitor the strain Md, which is dertncd as the temperature above on cooling under a constant load and the recovery which martensite cannot be stress-induced, may on heating. be about 25 to 50 °C (50 to 100 OF) higher than Af· Other important relationships of transforma- Ti-Ni Shape Memory Alloys I 1041 Strain of a Ti-Ni-Nb specimen M. temperature as a function of cold working Tempera t ... re. 'F ·300 -200 · 100 0 100 200 ,.r;~~~C--"~--~--~~--~ ~ ~ ~ " •& .;;.' 2 ..! , - -"- -2 " ,.L-______--J ·20 ·200 "00 o '00 20 25 30 Temperat ... re, ' C " " Cotd work, % " " Strain after delOfmlng and unloading, measured on tne first and Q1angeln lt1e M. temperatureol a Ti·SO.eNi ai oy cold worked 9.2 MCOrld h.ating cyd es. Note lMe cl"1ange in A, and 1M. rteOVfIry to 4004 and 5ubseq... ently anneated at 500 'C (9X1'F) tor 30 min. strain. SoI.Irot: G.A. Zact10 and T.W. o...erig, """p"'b'is.hed rI!$earch Scuret: K.N. Menon, J.L. Prell. and T.W. DtJerlg, MRS Int. M ee~ng on Actvanoed Materiats. \1019, K. Otsukaand K. SNmiru, Ed., Mat,. fIaIs Rtseartll Scciely, 1989. p 165 Ti·NI shape memory transformation temperatures Mel. which is defined as the temperature above which martensite cannot be stress induced, may be abo... tlrom 25 10 50 "C (SO to l00 "F) higher than At- Compa.ition Thn\perature. "C • ....t,':iN ; "- >f, A. T""'hniq .... ". ")7 " " '" ... D " " '" "'2 -" " ")) J2'" '" 0 -" ,,~ -'" " ~ -_lI'" .. ,,.. 21 -.." m [SIMcll .9.4 -", 03 "., " -lO '''' "'., _'"liO ->l ," 49.7 ., -"" IBOMiJI " '""'. , .. '''' "'., - 9._"2 9 " [68 W;anl 2O tol!i -" Elcc1ri.:al. ""'8n<:lic " n " "" fIRlPC "" [79OlcI 48.1 ' 00 ., .. , Eb:ui<:>.l ""i'livi!)' 48.6 '" 49.0 '".. " ". '" ,,~ .,, " ,"., " "'~ -"lO .." ". .:!! .., -•'" -J' Compilation UoIXI Ph.;..DiavamsofBiruu:- TIlaniu", A/kIyI, (J .L :'dW11l),. Ed.), ASM InternotionAl, 1987, p203. la )CiIed rcl.-~ &I'\! loS rouo ..... : 11 Kor. LL Kornilo¥, Yo:. V. Kachur, ...d O.K. &010\11.0", "Dilatation An.a]ys;,O!'1"ran'(omIltion in the CompoundTlNi: F iz. Md . Mrl41UwtJ. . 32(2), 42().422 (1971) in Ruuian; TR: Phyr. Mff. MnolkKr. • 32(21, 19I)..193 ( 197U. 81 l i e! : KX Melton and O. l lemer. "'The MechaNcal Propertiet o(NiTI.Bued Slape Memory AlIo)'t . ~ llcIo MtloU., 29. 393-398119811. 80 Mil: RY. l{;lIipn. "O.lf!rmination of Phue n.nsroro>ation 'ThlXlperat.u.resor TI.'1i U.mg Oilfef"eflti.a] Thmn.al An.a]y.i• •~1'i tani ... m"80. Ti Sci.1kh. . Proc. lnt. Con(. Kyoto, J". pM. M.yl8.22, T. Kiltluzi, Ed., 14S1-1467f1980J. 68W&n: F.£. Wang,B.r. OeSa"age,MrlW.J. But'hler. lllo: ' n-evtTSibl .. Critiaol RIInJ{f' ;" the Til'll n-..osltion: J . Appl. PII)'$ .• 39(5). 216&2115 (l9681. "19Che: D.B. Chlll"TlO'l', Yu .l. p"pal, V.E. G)'\lnter.LA ~ "",,sevio: h . "nd E.~1. lMvil-lkii. 1"he Mwtiplicity or Stroctunoll'raMil.ioaa iD A1 10)'t &5o:d on n:-o,: DoIIL lU ..d . NOIlIt. SS$R, 24;.854-851 ( 19191in It.... i . ..; 'I"R: Sot.r. Ph:#- DolL, 24{S I. 664-066(19i91 1042/ Advanced Materials Tensile Properties In general, a superelastic curve is charac terized by regions of nearly constant stress upon Schematic of superelasticity descriptors iooding(referre Ti-Ni shape memory: Typical loading and unloading characteristics l.oadi:Il pl:lllliIU 45OlO1OOMh UIIIoadin. pI=u Up.o 250 MP>. MIUimwn ~prtn8~k ,.. MWrnum defQrm;l.tiQI1 with % pcrmanc:m~ .. M.ulmwn lIOItd enon' " $our«: T.W. Dumng and C.R. Zadno. E~TlffriT16 Mptc1' of Slwtpe M~mcryNlay., T.W. Dueriget ...1., Ed., Butterwon..b.t!4m. mann, London. 1990, p 369 Strain SthemllIie dilgam ~ key dHcrIptors 01 superetastidty: 0u (l.WlIOIIcing jtateau mtasured aslhe inIe<:tionpoinl), o,(IoDIg pta. tQu rne&$Ul'1!d IS 1M ;"tIection point). (, (lota! deIofmation sIrai'I). Above M.;t ~ is defined as the temperature above which I, (permanent se~ or ..-mesiaJ and Ihe stored energy (shftded martensite cannot be stress-induced. Conse ~• . quently, titaniwn-nickel remains austenite Soutce: T.W. OIJerig tnd GR Zachl. Et>ginetriJg ~ cI throughout an entire tensile test above~. Tensile Shape Memcvy -/'So T.W. Duerig ef ai. Ed ~ 8uUeI ..OI1h-4Ieine strengths depend strongly on alloy condition, and mam. LoncIon. 1990. P 369 the ultimate tensile strength, yield strength, and ductility of cold worked titanium-nickel 'Ni.re de pend on final annealing temperatures (see figure). Aging of nickel-rich alloys increases llu!>terutic strength to a typical peak strength of800 MPa( 11 6 Ductility drops sharply as compositions be keD. Surprisingly, ductility is also increased duro come nickel-rich. A review of othec facton control ing the aging process. ling ductility can be found in Ref36. Mechanical properties vs anneal temperature Yield strength vs anneal temperature Temperature, OF Temperature. OF .,,, , 100 200 300 ,co 500 le'";800~,,~'~"~ __'"OCO~ __ C'~'''ec __ ~''~''~_ __ ' ~'~'' SO - , '" UTS r Ducti~ty- 'co '" I " • "' '" , •,; • , 109 • I "• "• & ~• l%YS I ~ ...... --- I ..~~.~.~--~j ~::::::::::::::::::J" 300 Theinlluenceol~*'9lemperal\.O'uonrned\arOcal~HoI Theinll..e-lced ll'f'ttoa6rlg~on/1"!fd"larblproperliesolli · 0.5 111m (0.02 In.) T~SO . 6Ni wire ..... '" 40% eok1 WOI1t and aMNlecI 50.6Ni ($j -..o1<.....-.eMd 30 ...... at 1emper.1IUu. 3Ominat~tur. . """40"'10 Scuroe: G.A. Zadno and T.W. Ouerig. t.npUbished re$eardl ~: G,A. Zact-o and T.W. Duerig.l..I"IpUbIished research BelowMs Titanium-nickel yield stresses are controlled to 13 \esi) foe titanium-nickel-copper alloys. Ulti by the "friction& of the martensite twin interfaces. male tensile strengths and ductilities nre similar Typical yields stresses are 120 to 160 MPa (17 to 23 to austenitic values. ksi) for binary alloye and as low as 60 to 90 MPa (9 Ti-Ni Shape Memory Alloys /1043 Superelasticlty (Between M. and Mdl Effect of Between M. and ~, the material transrorms shape memory and superelastic properties. Shupe Temperature from austenite to martensite during tensile test· memory and superelnticity per se wi ll not be reo ing. Yield strengths vary continuously from Ma to viewed; readers are rererred to ncf 1 to 3 for basic Md (see figure). The rate of stress increase is called information on titanium·nickel, Tee'MI', the stress rate, varying from 3 to 20 MParC, with Memorite™, Nitinol, Tine'nl, and FlexonTIoI . rates generally increasing with M•. These terms do not rerer to single alloys or alloy Superelasticity, or pseudoelasticity. is an en· compositions, but to a family of al loys with proper hanced elasticity when unloading between ~ and ties that greatly depend on exact compositional ~'1ct. The Md transition is generally defined as the make·up, processing history, and small ternary ad· temperature above which stress-induced marten· ditions. Each manufacturer has its own series or site can no longer be formed. On a stress·tempera· alloy designations and specifications within the ture graph, Md is the temperature where the stress "Ti.·Ni" range. begins to level off. A second compucation that readers must ac· Superelasticity is also highly temperature de· knowledge is that all properties change signifi· pendent (see figures). Changing alloy composition cantly at the transfonnation temperatures M Mr, and heat treatment can shift the temperature A., and Ar (see fib'Uro). Moreover, thege tempera" range of superelastic behavior from -100 to +100 tures depend on applied stres!. Thus any giv~n °C (-148 to 212 OF). property depends on temperature, stre!II, and his· This datasheet describes some of the key prop tory. Superclasticily is an enhanced elasticity oc· erties of equiatomic and near-equiatomic tita curring when unloading between AI and ?t'ld (see nium-nickel alloys with compositions yielding the section '!ensile Properties~ in thls da1asheet.l Permanent set 01 superelastic wire Loading and unloading plateau heights in Ti·Ni wire Temoera!ure. OF Tempe"lu'•. OF -300 ·200 ·100 0 100 200 300 .:)00 ·200 ·100 0 100 200 JOO " .- ~• , , ~ i" \ _~,oo[==:_,:oo=:::::-·o~-=-=,~oo::::_J,,;, -200 -'00 o '00 200 Temperat",re. "C Te-mper~lVra. "C Pt!Tn¥oent 5et 01 sup8felU!ic: binary titari~..we defotmld SO.8 at.'" Ni cold ~.., High-Temperature Behavior Stress relaxation ,.~------, o. Creep, Very few creep measurements have o. been made on titanium-nickel, although creep CJo " 564 MPa mechanisms have been proposed (Ref37), o. Stress relaxation is critical in many con o. strained recovery applications. Several measure ments have been made (Ref9, 12, 38). 'Ib summa i o. rize, relaxation of stable titanium-nickel alloys is 0.' extremely slow below 350 "C (660 "F) and becomes 0.' very rapid by 425 "C (795 oF). o 0.' o.. ~~~~~~~ __~~~ ·1.0 ·0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Jog (". h Stress rel&lCOlion 01 a Ni.rll",Fe) attoy at 375 ' e (705 ' F) measured ir1 a dynamieally oontroll&d Cteep mact1ioe. Round specimens YAIh gage lengIh 01 6 mm. hAy annealed. Source: J. Pro!! and T.W. Duerlg. Engineering Aspet;1s 01 SI!~ Memory AIo}'$. T.W. Ouerig e/ aI., Ed .• Butte Fatigue Properties The fatigue behavior of titanium-nickel alloys controlled environments. The superclastic meehl!. is extremely complex and encompasses m8J)Y dif· msms accommodate high strains without exces ferent topics. 'Ib summarize, titanium-nickel ex sive stresses, but cannot accommodate high cels in low-cycle, strain-controUed environments, stresses wi thout excessively high strains. and does relatively poorly in high-cycle, stress· Stress Isothennal stress-controlled testing is impor sively characterized by the coupling and fastener Controffed tant in many constrained recovery applications in industries, much of the data remains largely un which the shape memory effect is only \;sed as a published or highly application specific. Typical S· mode of installation, e.g., fasteners and couplings, N behavior for vruious alloy compositions tested in which the nlloys are used exclusively Ilhove the weH above lheir Md temperatures is shown below Md temperature. Although fatigue has been exten- (see figure on next pnge). Strain Isothermal strain-controiled behavior is highly lar interest is superelastic cycling (strain-control· Controlled dependent on the testing temperature relative to led testing between AI and l'v1.!: ). There are severnl the alloy transformation temperature, Fracture modes of degradation that occur under these cir· follows the Coffin·Manson relationship: cumsto.nces (see figure), Again, these depend NI' • .,,=C strongly on specific alloy conditions and the needs where C and p areconstants;N is the number ofcy of the application, Further data arc provided in cles to failure; and 6£p is the applied strain ampli Ref39. tude (see figure). One specific example of particu- Thermal Thermal cycling both with and without applied ing methods, heating and cooling rates, and even Fatigue loads such a s found in actuators or heat engines, is orientation (hori:tontal or vertical I. certainly the most complex mode to analyze. Al Failure too is nebulous. Various degradation though it has been studied extensively. it remains modes can occur, including a shift. in transforma' difficult to predict failW'e, largely because failure tion temperature, a reduction in the avnilablc can consist of fracture, ratcheting, migTation of strain, walking (o r rntcheting), and fracture itself. transformation temperatures, changes in reset Thermo.! cycling with no applied lood can even re force, etc. Moreover, damage accumulation de sult in some degradation. See Ref 10 to 13 for fur pends on stress, strain, temperature change, heat- ther information. Ti·Ni Shape Memory Alloys 11045 Typical S·N curves Low.c:ycle fatigue behavior ".""------..., , M.2.70 · C • M, .. .. 7.S · C • M..,·oo~ 0 M.,,·t20~ • , Electron beam melting ,.".~::;~t::-,,:~~~:.~-, ~ II = 0.155 o T;'5O.5 Ni ... . ~10" o T .. 5O.7 Ni , j} ,,0. 176 Ti·5O.9 Ni • ~• (l &11.216 • r..so.& Ni High--lrequency • T;'51.0 Ni vacuum induction melting j},,0.167 -30,,'L__ ~_~~_~:__- ...J 5 10·tL_ _ ~ ___,-;- _ _ ...,:-_-1 101 10) 10' 10 10' 10 lot 103 10' Cycles 10 failure Cycles to failure titanium-niCkel aIToys of varioo.s CO'TIposiUons (at. ~.) prepared In varioos fashionsl! lsothermaJlytested in sI1eet form. Tes~ng is done l.J:lw-.<:yOe lallgue bMal'lor of titanium·nlckel alloys at room tern· well above M~. SMeal specimens loaded and unloaded (R = 0). Cy· perature testGCI in tllnsion-eompression (R ", -1). The M, tem pera· de-dat 6O · C (140 · ~ . turM 01 the alloys are ill(iiQJted. Source: S. Miyazaki. Y. Sugaya. and K. Otsuka . MRS 1,,1. MeeUng SOurce: K.N. Melton and O. Mercier. SIrength of Marais and Alloys. on Adv.moed Male<'lals, Vd 9. 1(. Otsu Effects of isothennal superelastic cycling • -2 , • • 12°C • • , • 26 °c • , , , , • • , " ·C , , .. • , , , , , , , , .. 4(FO,--GN ,•, • • 2 '. • 0 > , 0 o • '. •, • g- • • • • • • < " • • • • 12°C • " • • ,,'C , • • • o 0 0 0 •0 .. ,.'C " ,. • ••• , '00 , 25 '00 " Cycles to failure '" " Cycles 10 railllr~ '" '" (.) (b) ,,------, ~ooo 0 .... 0 0 • o • • • • • • , 12 • • • , , . . • • • , • • • • • • 0 • o • • t - • , 50 "" Cycles to 'a~ure '" (.) THIs 01 a superelastic ~tamm~ "'r8 at Itvee tempeBtures. AJloyccnl:illf\$ 50.6 at.% noc ~e.. Throe modes of deg'adalion oa:ur' sirooItan& ously do..oing Itre isolhennal superelastic eycIing 01 titanio..nl·nidte! alloys: v.9I<.ing. 01 an aocumulallon 01 permanent Sol! (lOp). a chaoge in yield s!JeSS (middle) and a reduction in :tle hysteresis wiaIrr (bottom). Source: S . Miyazaki, EngiIleemgAspecfsolShapeMemotyAlloy$, T.W. Ouerig ~ aI., Ed., BuHerworths, 1990. p400 ,0461 Advanced Materials Fracture Fatigue Crack Strcss-inducing martensite at the tip of a slow crack growth (Ref 44). Othcr sources ofdala Propagation propagating crack has been shown to significantly include Ref 45. Toughness Although Charpy impact testing on titaniwn sharp toughness minimum exist.OJ at Md (see fig· nickel has been conducted (Ref3, 46). very little Kfc ure). or JIc data exist. Indications are strong that a Fatigue crack propagation ,,' 0 Stable marten$ite • Stable austenite • Irreversible stress-induced mallMsite ,,' , Reversible stress·induced marteMite i.! 10"J r i,,{ One laltice spaOng eer eycI.!... __"""" • • 310 "~ • o • • 10'''[ ,,. Slless-itU:ensity range" (,6,,1(). MPa,:m FaSg\:e crac;Ii: ptepaga:ion ra:esin 1000titanit.m.ndlel iIIIo)'$.(~ staJ:H~~le"';:hM. > tOoI'\'I ~tufe. stabl6'3USkYIiIe"';lh M" « room tempetalu'e. ~ stress~manMSiIiJ-M!h M. < room tempefalVrc < I\- and ~e sR$.$-it)ducedmar/ensi9..,;th 1\ < roorn:~1\.We Processing Fracture energy vs temperature Flow Stress. Upset forging tests conducted Bulk Working on Temperature, ' F seven different alloys at strain rates ranging from 2 , '" 200 10~ to 10 indicate that the material has a very OS high hot ductility and is not highly strain rate de pendent. The £low stresses can be characterized by • 0,30 the relation: •E 0.25 a=kr.m exp(-Q1R11 ~ .'" " where a is the flow stress; r. is the strain rate; Q is an Fracture energv ! activation energy (50,000 cal/mole); and m andk are • constants. Values for m were found to range from 0.1 ." • to 0.17 with increasing temperature (see figure for typical data ofan equiatomic binary alloy ). " J OS Extrusion. Both solid bar and large tube (50 ~-.-' - Yiekl ~rength (0.2%) mm mameter with 8 mm wall truckneB$) have been 00 extruded from titanium-nickel. Parameters re ·eo , eo main proprietary. Temperature. 'C '''' Forging. Titanium-nickel has been succes Vac:u!.m ~ meiled alloy hot VoOI1Ied and <'Mea!ed lor I h;l.1 fully forged into large cups. Parameters remain 950 "C and aircooled. SlandardCha'll'lV.1'IOId1 impacl speeirnen$ proprietary. were mac:hine Fabrication Casting. Although some unreported experi· Flow stress measurements ment:i have been conducted, casting has not been done on a commercial leveL Powder Metallurgy. Although a great deal of • experimentation has taken place wit h both ele· • V1 8ls ~ mental and prealloyed powders, nothing hns , reached near·production levels. Refe rence 5 pro -- 0.11315 vides a review of methods. ~ > 5. Forming. Titanium-nickel sheet has been suc o cessfully fonned into a range of complex shapes, ; ;;: 5. O.OIZ3Is both in the martensite and austenitic phases. S Springback is high, as is die wear and friction. Pa· rameters remain proprietary. Machining. Titanium-nickel is very difficult .. to machine. Very low speeds and a great deal of • coolant is required, and tool wear is very rapid. ' .5,'-"-______-' Milling and drilling are particularly difficult. Pro 0.' 0.' '.0 ducers of couplings have demonstrated that large lfT x 1000" " " scole machining production is possible. Nal~(81Iog (In) al fiow stress is shown la var, l inea ~y with the in· Heat Treatment. Titonium·nickel can be heat Yellie al abSOllile temperalvre. 12. mm x 8 mm diameler specimens treated in air up to -500 "C (930 "F). No 0: case is lesfed In oompres.si()n I.nder isothermal condilKlf1S (healed dies) at formed, but a surface oxide of rutile develops !hit I!!mp&ralvres and slraon ra les shown. SOvree: T.W. Dvttti9. unputllisheddala quickly. Above 500 °C (930 oF), the oxide layer be· gins to flake (depending on time). Nitrogen and hy drogen at mospheres are not recommended. Argon, Joining. Titanium·nickel is difficult tojoin be· helium, and vacuum heat treatments are com· cause most mating materiuls cannol tolerate the monly used to preserve bright finishes. large strains experienced by the alloy. Most appli Recrystallization is ext remely rapid above 700 cations rely on crimped bonds. It can be welded to °C (1290 OF ). Solution treatment requires tern· itself with relative ease by resistance and TIC peratures of at least 550 "C (1020 "F). Stress relief methods. Welding to other materials is ext remely is usually accomplished at temperatures as low as difficult. although proprietary methods do exist 300 °C (570 oF). A Tn' diagram is shown in the and are practiced in large volumes in lhe produc. previous section "Phases and Structures. ~ tion of eyeglass frames. Fully annealed bar stock has a typical hard Brazing can only be accomplished after re-en' ness of 60 HRA. Vicker numbers range from 190 forcement and plating. Again, methods are pro· HV in the annealed condition to 240 HV after 15% prietary; large-scale production is practiced by the cold work (Ref3). eyeglass frame industry. References . 1. T.W. Duerig et al.. Ed .. Engineering Aspects of London, 1990, P 181 Shape Memory Alloys, Butterworth-Heine· 11. T.W. Dueng ami G.R. Zadno. Ellyincl'nnl! A~ · maM, London, 1990 peets of Shape JJemory A ll())".~. T.w. Duerig cl 2. J. Perkins, Ed.. Shape Memory Effects in Al al.. Ed., Butterworth-Heinemnnn, London, lay.';;, Plenum Press, 1975 1990, p 369 3. C:M. Jackson. H.J. Wagner, and RJ. 12. C.R Such. 'The Charnctenmtion ofthe Rever Wasilewski, NASA Report, :-iASA·SP 5110, sion Stress in XiTi," M.S. Thesis. Naval Post 1972 Graduate School. Monterrev. CA. 1974 4. J. Takahashi, :\1 Oka;reki, K Hiroshi, and Y. 13. M.V. r>:evitt. Thln.~. Met. SOC. AlME, Vol 218, FUIllta. private co mmunication. 1984 1960, p327 5. T.W. Duerig, in Aduan.ced S_vnthesis of Materi· 14. D.M. Goldstein, W.J. Buehler. and R.C. Wiley, als, J. Moore, Ed., 1993, to be published Effects of Al loying upon Certain Properties of 6. G.R Zadno and T.w. Duerig. Engineering As· 55.1 Nitinol, NOLTR 64·235, 1964 pects of Shape .\femory Alloys. T.W. Duerig et 15. H.K Eckelmever. Sandia Labs Report 74- ai.. Buttenvorth-Heinemann. London, 1990, p 04 18.1974 . 4 14 16. D.D. Chemovet ai.. Dohl. Aked. ,va.uk. SSSR , 7. M. Nishida, C.M. Wayman and T. Honma, Met· Vol 245 (:'-io. 2). 1979, p 360 all. 1rans. A , Vol 17, 1986, p 1505 17. P.G. Lindquist and C.M. Wayman. Enginccr. 8. TW Duerig and KN. :Vlelton,?roc. ofSMA'86, ing Aspects of Shape Mem,:.-y Alloys. TW. C. Youyi et ai., Ed., China Academic Publish Dueng et al.. Butt.cI"Worth-Heincmann, Lon· ers. 1986, p 397 don, 1990, p 58 9. J.L. Prott and T.W. Duerig, Engineering As· 18. T.w. Dueng and KN. Melton. The Marten.~i!c pects of Shape Memory Alloys. T.W Duerig et Transformu.ti!lTl in Science and Technology, E. al., Butterworth·Heinemann. London, 1990, p H om~l .!I en und H. Jost, Ed., Deutsche Gesell· 115 schall. :l..!r !Iiolclallkunde, Gcrmuny. 1988, p 191 10. A. Keeley, D. StOckel, and T.\V. Dueng, Engi 19. W. MO!H.! rly and K. Melton, Engineering A~ · neering Aspect.~ of Shape .vfemory Alloys, T.W. pcc:l .~ ot" S hnpe .\femllry Alloys. T \V. Duorig et Duorig et of., Ed., Butterworth-Heinemann, 1048/ Advanced Materials ai., Ed., Butterworth.Heinemann, London, C.M. Wayman, Metail. '1h:Jns., Vol 2, 1971, P 1990,p46 2583 ZO. O. Matsumoto, S. Miyazaki, K Otsuka, and H. 36. S. Miyazakiet al., Mater:. Sci. Fomm, Vol 56-58, Tamura,Acta Metall., VollS, 1987, p 2137 1990,p765 21. H.C. Li.ogand R. Kaplow,Metall. ThJn$. A, Vol 37. A.K MukhcJjee, J . Appl. Phys., Vol 39 (No. 5), 12,1981, p 2101 1968.p2201 22. L. Kaufman, SA Kulin, P. Neshe, and R 38. A. Thrui et ai., Nip. Kihai Gakhai ROllbunshu, Salzbrenner. ShaIN Memory Effects in Alloys, A Hen, Vol 51 (Nu. 462), 1985, p 488 J. Perkins, Ed.., Plenum Press, 1975, p 547 39. G.R Zadno. W. Yu, and T.W. Duerig. MaU!rials 23. S. Spinner and AG. Hozner, J. Acoust. Soc. Scknce Forum, Vo l 56-58, D.C. Muddle, Ed., Am., Vol 40, 1966, P 1009 1990, p771 24. R.J. Wasilewski, 'Ihm.s. AIME, Vol 233, 1965, p 40. H. Tamura, Y. Suzuki and T. 'Ibdoroki, Prot. 1691 lot. Con£. MartensiticTro.nsrormations, Japan 25. O.K Belousov,Russ. Metall.. Vo12, 1981, p 204 Institute orMetals, 1986, p 736 26. J .E. Hanlon, S.R. Butler, andR.J. Wasilewski, 41. Y. Furuya et ai., MRS Int. Meeting on Ad ']}ans. AlME, Vol 239, 1967, p 1323 vanced Materials, Vol 9, K. Otsuka and K. 27. CoM. Chan, S. Trigwell, and T.W. Duerig, Surf. Shimizu. Ed., Materials Research Society, InterfaceAncl. Vol IS, 1990, p 349 1989, p269 28. J.D. Harrison, ~ychem Report, 1991 42. Y. Suzuki and H. Tamura, Engineering A<;pect.<; 29. S. Lu, Engineering Aspects of Shape Memory ofShapeMemoryA1.10ys, T.w. Dueriget 0.1. , Ed., Alloys, T.W. DuerigetaI., Ed .. Butterworth-He l3utterworth-Heinemann, London, 1990, p 256 inemann. London, 1990, p 445 43. J.L. Pratt, K.N. Melton, and T.w. Duerig, MRS 30. L.S. Castleman, S.M. Motzkin. F.P. Alicandri, Int. Meeting on Advanced Materials, Vol 9, K VL. Bonawit, and AA Johnson, J. Biomed. Otsuka and K Shimizu. Ed., Materials Re Mater. Res., Vol 10, 1976, p 695 search Society, 1989, p 159 31. R Burch and N.B. Mason, J.e.s. FarcuJ.ay I, 44. R.H. Dauskardt, TW. Duerig.and RO. Richie. Vol 75, 1979, p 561; RB. Burch and N.B. Ma Jl..{RS Int. Meeting on Advanced Materials, Vol son,J.e .S. Faraday I, Vol 75. 1979, p578 9, K Otsuka and K Shimizu. Ed., Materials 32. AR Pelton and TW. Duerig, "'AIl Analysis of Resetl.n:h Society, 1989, p251 Crofit Couplings After One·Year Service at 45. S. Miyazaki. Y. Sugaya, a.nd K. Otsuka, MRS Seabrook,8 Raychem Proprietary Report, 1991 lnt. Meeting on Advanced Materinls, Vol 9, K 33. G. Ai.roldi and B.Rivolta,Phys. Scn'pto , Vol 37, Otsuka and K Shimizu, Ed. . Materials He 1988, P 891 search Society. 1989, p 251 34. J.C. Gachon and J. Hertz, CALPHAD, Vol 7, 46. K N. Melton and O. Mercier, Acta Metall., Vol 1983, P 1 29, 1982, p 393 35. K Otsuka, T. Sawamura, K Shimizu, and