Influence of Ultrasonic-Shot Peening on Bending Fatigue of TiNi Shape Memory Alloy

(1)

第 17号 2015年

I

n

f

l

u

e

n

c

e

o

f

U

l

t

r

a

s

o

n

i

c

-

S

h

o

t

P

e

n

i

n

g

on B

e

n

d

i

n

g

F

a

t

i

g

u

e

o

f

T

i

N

i

S

h

a

p

e

Memoη

T

A

l

o

y

Kohei Takeda

t

，

Ryosuke M

a

t

s

u

i

t

a

n

d

H

i

s

a

k

i

T

o

b

u

s

h

i

t

Abstract The fatigue property of shape memory alloy (SMA) is one ofthe most important su対ectsin view of evaluating functional characteristics of S1ι'¥ elements. In the present study， ultrasonic shot peening (USP) was applied to induce compr巳ssiveresidual stress on the surface layer ofTiNi SMA tape and the influence ofUSP

on the bending fatigue life was investigated. Th巴 fatiguelife of USP-treated tape is longer than出atof the as-received tape. The fati忽Ielife of也etape USP-treated with high coverage is longer than that with low coverag巴.The fatigue life of the USP-treated tape increases in proportion to the hardness on th巴surfaceof the

tape.

1. Introduction

The shap巴memoryalloy (SMA) is expected to be applied as

intelligent material since it shows theUllique characteristics of the shape memory effect and super巴lasticity1).The TiNi SMAs

紅巴 amongthe best SMAs since the fatigue life is longer，也e recov己:rablestrain is larger and the corrosion resistance is higher than0出巴rS恥1As.With respect to the fatigue prop出i巴民社le fatigue tests in tension and rotating bending have been conducted2-4).However，也eplan巴四bendingdeformation is used in practical applications.Inthe growing number of TiNi SMA applications，仕lesematerials should fulfill high陀午lIrementsof fatigueラcorrosionand we訂 resistance.On the other h釦d，the application of SMA has some limitations， particularly in thermomechanical cyclic loading cases. If the maximum strain is I紅ge，mechanical property c釦 ch釦geduring cycling and S仕切turalcomponents c組 bedamaged dueωfatigue under high cycles5-6).Intheseωses， fatigue of SML屯isone of也巴出portant properties in view of evaluating functional characteristics as SMA elements. It is known也at the shot peening is used to induce compressive residual s仕'esseson the surface layer of metallic partsη. The r巳sidualstresses produced increase the fatigu巴 performances. The ultrasonic shot p官邸ling (USP) has th巴

t

Depar也lentof Mechanical EngineeringラAichiInsti加teof Technology following advantages: (1) saving space since cabinet and dust collector訂enot needed， (2) clean operation lines with improved operation environment (dust and noise) because of no dust collector used in lines and (3) shot peening effect with the same or higher

∞

mpressive residual s仕esscompared to conventional shot peening and with the betler surface fmishing血an conventional shot pee凶ng8).It is suggested that USP would produce SMA elements with a longer working life. The study on enhancement of也efatigue life for T山iSMA by USP has not been investigated till now. In p訂ticular，sin

∞

impressions appear little by USP in出巴 region of superelasticity，註le influence ofUSP on the fatigue life is not cl巴飢 In this paper， we report and discuss the influence of USP shot media diameter and coverage on the bending fatigu巴lifeof

T山iSMA tape which shows superelasticity.

2. Experimental Procedure

The mat巴rialsused in the experiment were Ti・50.85at%Ni

tapes with a thickness of 1 m m and a width of 2.5 mm. The length of批 tapeswas 100 m m for the tension回tand 80 m m

for the fatigu巴test.They were polycrystallineヲproducedby

Furukawa Techno Material Co.， Ltd. and applied to an element of a brassiere. Th巴 tapeswere heat-仕官atedat 803 K for 10 min

in air followed by water quench.The material shows superelasticity at room temperature.

(2)

84 愛知工業大学総合技術研究所研究報告，第 17号， 2015年 In 'Order t'Oinvestigate吐leinflu巴nce'Of USP 'On the fatigue life， the tapes were USP-甘'eated.In 'Order t'Oinvestigat巴theinfluence 'Of surface roughness'On the fatigue life，出esurface'Of the tape in 'One gr'Oup was p'Olished by using the abrasive paper N'O1000ラ 1200， 1500，2000，2500 and吐lenfinished using 0.04 micr'On Si02 c'Oll'Oida1 suspensi'On. The specimen were divided int'Othe f'Oll'Owing gr'Oups: (1) as-received， (2) p'Olished and (3) three kinds'OfUSP屯eatedtapes

On巴巴nd'Of a TiNi SMA tape was gripped during USP

trea位len.Bt 'Oth flat surfaces'Of the tape were sh'Ot-peened企om れ;V'O 'Opp'Osite directi'Ons by st巴elballs as sh'Ot media.The sh'Ot media was an SUJ2 steel ball with Vickers hardness HV 'Of 850 used f'Or a miniature beぽmg.Diameters'Of the steel balls used were 0.8 mm and1.2 mm. USP was c紅ried'Out using th巴 ultras'Onic equipment produced by SONATS (T'Oy'OSeik'OC'O.， Ltd.).The企己quency'Of the vibrating s'On'O仕'Od巴was20魁Iz.The impressi'Ons d'On'Ot clearly appear'On the sh'Ot-peened surface'Of 也esuperelastic SMA. In 'Order t'Ospeci今出epeening intensi句I'Of

USPラAhn巴ntest strips wer官used.The c'Overage was 'Obtainedお

おll'Ows:(1)白etime'Ofthe USP-仕切削lentf'Or Ahnen test s汀ipst'O the

∞

verage'Of 100% wぉmeasuredat frrst， and (2)ヲin吐lec蹴 'Of c'Overage'Of 2000%，出eUSP-treatrnent was applied t'O血e SMA tap巴f'Or20 times l'Onger time血組that'Of吐lec'Overage'Of 100%. C'Overages applied were 2000% and 4000%. In也etest t'O investigate the influence'Of the sh'Ot media diameter'On the fatigue pr'Operties， tw'Odiameters'Of 0.8 m m and1.2 m m w巴reused f'Or a c'Overage'Of 2000%. In the test t'Odiscuss the influence'Of也巴 c'Overage， tw'Oc'Overages'Of 2000% and 4000% were applied f'Or a sh'Ot media diameter'Of 0.8 m m In the t巴nsi'On test， displacement was measured by an extens'Ometer with a gauge 1四割1'Of 50 m m f'Or SMA tapes

Straint:was determined by n'Omina1 s廿ain.The t巴nsi'Ontest was

carried'Out under a c'Onstant s廿ainrate dt: / dt = 1.67

x

10-4 S-1in

mr at r'O'Om tempera旬reTr

I n 血efatiglle testヲ組altemating-p lml巴bendingfatigue test

machine9)was used. The test f'Or altemating-plane bending

fatigue was c紅ried'Out in air at r'O'Om temperature. The maximllm

bending s仕ainappeared'On the flat surface'Of the SMA tape at

which USP was applied. The frequency'Of cyclic bending wぉ

f'Ound t'Obe 2.5 Hz (150 cpm). Temperature'Of the specimen surface at the midp'Oint'Oftw'Ogrips， where the maximum bending strain appeare，dwas measured by a therm'Oc'Ollple during the altemating胴planebending fatigue test. A scarming electr'On micr'Osc'Ope (SEM) was usedω'Observe the frac旬resurfac巴'Ofthespecimen.

A laser microsc'Ope VK-X200 made by Key巴nceC'O.was used

t'O 'Observe the surface r'Oughness'Ofthe specimen. 百leVickers h紅白esstest machine was used t'O 'Observe the surfac巴hardness'Ofthe specimen. The indentati'On l'Oad was 49 N. The residual stress was measured by Xィaydif企acti'Onusing 28-sin2_V_'_d_i_a_g_r_a_m_s_f_'_O_r_t_h_e₍₂₁₁₎_p_l_a_n_e_'_O_f_白_ea_u_s_t_e_凶_t_ep_h_a_s_e_. 3. Results and Discussion 3.1 Tensile d.

φ

rmation property 百lestress-s廿aincurves'Of f'Our記nds'Of SMA tapes'Obtained by the tensi'On test are sh'Own in Fig.1.All stress-s甘aincurves draw hysteresis l'O'Ops during l'Oading andunl'Oading， sh'Owing th巴 superelasticity. The upper stress plateall during l'Oading appears due t'O也estress-induced martensitic transf'Ormati'On (SIMT). In 白巴S仕'essplateau regi'On，出etransf'Ormati'On band progresses due

t'Othe SIMT in TiNi SMA tape10)The l'Ower stress plateau

during unl'Oading appears due t'Othe r巴vers巳仕組sf'Ormati'On.The

martensitic甘ansf'Ormati'Onst訂tstress句1Swas 'Obtained企'Omthe

intersecti'On 'Of tw'Ostraight lines: th巴initialelastic p釘tand吐le upper stress plateau p釘t.Va1ues'Of句1Sare 370 MPa， 3801-住民 400 MPa and 410 MPa f'Or the aトreceivedtape，也巴 tape USP・treatedwith sh'Ot media diam巴terd = 0.8 m m and c'Overage c =2000%ラ白atwithd = 1.2 m m and c = 2000%，組d白atwiせld= 0.8 m m and c = 4000%， respectiv巴ly.The value'Ofσ'Ms mcreases slightly by USP since the influence'Of USP appears'Only'On the sh'Ot-peened surfac巴layer'Of the tape. Al也'Oughb'Oth the sh'Ot media diameter d and c'Overage c affect the value句必 the 凶 uence'Of c'Over噌 cis a little higherせlanthほ'Ofsh'Ot media diamet巴rd. 事#。骨骨

z

4 dddt~芯1.岳7x 1詰"'ft・5 T， ~2W， ~29唱 K 品 St問inc(弘 } Fig. 1 Stress-strain curves 事

(3)

0.8 m m at c= 0.52 and 0.41 for bo也tapesUSP-treated with d 2000% and with d = 1.2 mm at c = 2000%ラand0.37 and 0.33 for 3.2 Bending Fatigue Property 出巴tapeUSP-tr巴atedwith d = 0.8 mm at c = 4000%ラrespectively. The calculated results ofEq.(1) are shown by solids lines in Fig. 2. The overall inclinations are well approximated by the solid (2) Temperature rise As observed in Fig. 2， the larger白巴bendingstrain amplitude，白eshorter the fatigue life is. During cyclic bendingラSMAis subjected to巴xo仕1ermicmartensitic transformation repeatedly. Therefore， temperature of the material increases markedly in the early 20 -30 cycles and is saturated泊acertain value thereafter

ll)百四di日ferencebetween the saturated temperatureTs and room

tempera旬reTr is defmed as the saturated tempera加rerise !1T

，

=

Ts -T，..The relationships beれ同enthe sa知ratedtemperature rise !1Ts and the bending s仕ainampliωde &a for four kinds of臼p巴S

obtained from the test at rロomtemperatureT，.are shown in Fig.3. InFig.3，吐1edata obtained紅巳plo臨dby several symbols for each tape.Th巴 region，where the data are plotted， is shaded. Temperature rise !1Ts increases泊proportionto&a for all tapes and the difference of !1Ts is slight翻 ongfour tapes. As seen in Fig.1， the stress-s仕aincurve of位1etap己drawsa hysteresis loop during loading and unloading. The area surrounded by出e hysteresis loop co町'esponds to the dissipated work per unit volume Wd 12).百1edissipated work

シ

T

f

is evaluated by the product of the diffi巴renc巴(σM一句)between the upper plateau S出ss句1and也elower plateau s廿ess句 and也巴 martensitic 回nsformationstrain range&M-百四 dissipatedwork Wd of也E surface element with the bending strain amplitude&a is estinmted by the following equation 13) lines. 3.2.1 F，αtigue 1ザ巴 (1) F，αtigue life curve

Th巴r巴lationshipsbetween the bending strain包npli知de&a and

批 nmberof cycles to failureN_ffor fiv巴凶dsof tapes obtained by the altemating-plane bending fatigue test under a constant 企equency

f=

150 cpm at room tempera加古閑shownin Fig. 2.

The bending strain amplitude&a was obtained from也己bending strain on the s町faceof出巴specimenat the合ac印repoint.T he specimen was fractured at the midpoint of two grips. As can be seen in Fig. 2ヲthelarger the bending strain ampliωde， the shorter the fatigue life is. The fatigue life of polished and shot-peened tapes is longer than that of白eas-received tape.百1efatigue life of the tape USP-treated withd

=

0.8 m m and c

=

2000% is almost the same as that withd = 1.2 m m and c = 2000%. The fatigue life of也etape USP-treated withd = 0.8 m m and c = 4000% is longer than that of oth巴'fshot岨peen巴dtapes.百1efatigu巴lifeof出etape

USP-treated withd

=

0.8 m m and c

=

4000% is 2.6 times at&a =

3% and 10 times at&a = 1% longer社1an仕1atofぉ-r巴ceivedtapeラ

r巴spectively. η1巴relationshipsbetween the bending strain amplitude&a and 出enmber of cycles to failure

め

shownon the logarithmic graph are almost巴xpressedby straight lines for all materials.The relationship therefore c加 beexpressed by a power function as ) I ( whereα組dβrepresentゐinNf= 1 and th巴slopeof the log &a-log

め

curve，respectively. Th己 valuesofαand

s

are1.16釦d 0.55 for theお開receivedtap巴，1.68加d0.55 for the polished tap，巴 follows: &a'N

/=

α

(2)

l

=

1

緊1cp描 As・g叫ci同d l'SP ~仏軍 X2晶君詣叱~. むSPOO.8X40明。呪. FSPo1.2X:2事昔時吃当 Wd=(σiI1-dA)(ε白-&MS) 1，事3111K己主怠2哲9 K o 冒 Z骨6KO = 294 K b. 111 与d

.

勺昌、包，拶宅 1:

g

品芝

き

4

‘

'

宅2 毛，

E2

:1

.

，

語

。

酵 1 2 3 4 B開din耳品開i露aml'lil認dむら(%) ltJh 10' 11ド 10タ h励。of{"wlcs帥おiI附吟 Fig. 3 Relationship between saturated temperature rise L1Ts and b巴ndings仕ainamplitudeら・ Fig. 2 Relationship beれveenbending strain釘nplitudeand the number of cycles to failure.

(4)

86 愛知工業大学総合技術研究所研究報告，第 17号， 2015年 Fig. 4 SEM photograph of企acturesurface of as司receivedtape for ゐ =1.26%釦dll

ケ

=5906 where 5MS denotes the martensitic transformation start s住ain. Therefore， if5a is 1紅ge，Wd is 1紅ge，組das a result加geamount of heat is generated. This is仕lereason why !J.Ts increases in proportion to んIn SMAs，σ'Mincreases in proportion to temperature as follows: d句1/ dT = CM and~'VI = 6 MPa/K for TiNi SMA 1の.lt should be noticed仕latヲif!J.Ts is large， (YM increases. Therefore， in白e practical application of SMA elements， hi俳句1induced due to temperaωre rise A

乙

under cyclic bending causes large fatigue damagふthatis，せlefatigue crack nucleates earlier and社le fatigue crack propagate rate becomes higher， resulting in short fatigue life. 3.2.2 Fatigue crack (1)FatigueSUiゆce Inthe cおeof as-rl巴C巴ivedtapes， the fatigu巴crackinitiates at thec印刷p制 of也eflat surface su吋ectedto maximum bending s仕ain.In仕lecase of USP tapesラthefatigue crack initiates at白e

comer different from the flat surface subjected to m似imum

bending strain. Let us discuss白einfluence ofUSP on the fatigu巴 crack initiation and出efatigue crack growth by observing the fractur巴SI町face. (i)As-received material Figure 4 shows the SEM photograph of a合ac同resurface of the as-received tape obtain巳dby the fatigue tl巴stfor5a = 1.26 %. ln Fig. 4， Fcdenotes吐lepoint of the fatigue crack initiation.Tbe crack nucleates at a certain pointFcin the central p紅tof白eflat surface of the tape and propagates toward the center in a sinuous radial pa仕em.Although small cracks are observed in both flat surfaces of the tape subjected to maximum bending strain， one single crack grows p肉食rentially.Following the appearance of fatigue crack with a semi-elliptical surface， unstable企ac加re inally occurs. In白ecase of as司.receivedtap巴s，也epoint 凡 appears at a c巴ntralp紅tof the flat sぽfacesubjected to th巴 maximum bending s仕泊nand吐lefatigue life is short (ii)USP material

SEM photographs of a企ac加resurfaωfor也巴USPtape in the

case of a bending s汀ainampli旬命令 =1.64%紅巳shownin Fig.

5.Tbe tape was USP-treated with shot media diameterd = 0.8 m m and

∞

Iveragec = 4000%. The whole企acturesurface，吐le fracture surface at the crack initiation part (1)ヲ白c合acturesurface in the middle part ofunstable fracture region (2)，加dthe企acture surface in the final unstable frac切rep訂t(3) located on仕le opposite side of白 crackinitiation p副 (1)紅白shownin Figs. 5 (a)，5(b)ラ5(c)姐d5 (d)ヲrespectively. As can be seen in Figs. 5 (a) and 5 (b)，也efatigue crack nucleates at a certain point凡in也ecomer (1) on社leside surface near the flat surface of the tape and propagates toward社lecenter in a sinuous radial pa杖.em.百四distance from the flat surface of the tape to the crack initiation point

三

}

is40戸n.Although small cracks are observed on four comer surfaces ofthe tape， one single crack grows preferentially.Tbe reason why the fatigue crack nucleates at the comer surface is as follows. Al也ough the maximum bending s仕ainappears on the flat surface of出巴tap巴ヲ the flat surface is su句ectedto USP and仕lereforeit is hard for the fatigue crack to nucleat巴on也巴flatsurface.Tbe side surface of the tape is subj巴ctedto sligbt USP. As a result， the fatigue crack

nucleates at the comer 九near也eflat surfac巴.百lIsphenomenon

is similar to the fatigue crack initiation point of a TilサiSMA wire subjected to nitrogen-ion implantation15) In血巴 case of飢

ion-implanted wireラthefatigue crack nucl巴atesat a certa泊point

different企om仕1巴maximumbending s甘ainpoint where the

maximum amount of nitrogen ion was implant巴d.Inpractical

applications of SMA丸thefatigue life of SMA elements increases if USP is treated not only on社lesurface at the maximum stress point but also on the surfa

∞

in the r巴gionnear仕lemaximum 5仕esspomt As can be seen in Fig. 5 (a)ラfollowingthe appearance of fatigue企acturewith a quarter.網ellipticalsurface， unstabl巴企acture finally occurs. As c釦 beseen in Fig. 5 (c)， isometric dimples with an average diameter of about 3戸nappear in the middle p紅t ofuns匂blefrac附 region(2). As seen in Fig. 5 (d)， elongat己d d出pl巴sar巴founddistributed in the direction parallel to也eflat surface according to the propagation of the crack induced by tearing in the final unstable fracture part (3).

(5)

(i) Fatiguecγackgrowthaγ巴α Let us discuss the fatigue crack growth area Aj which is surrounded by dashed semi-ell取icalor quarter-elliptical lines shown in Figs. 4 or 5(a).百1巴relationshipsb巴れ万巴en出巴fatigue crack growth areaAjand th巴bendingstrain釦lpliωde5a for four kinds of tapes obtained by也巴fatiguet巴，stare shown in Fig. 6. In Fig.6ラthedata obtained are plo仕edby several symbols for each tape 百1e regions， where th巴data are plo仕巴dラ are shaded separately for the USP tape withd

=

0.8 m m at c

=

4000 % and for other three tapes. As can be seen in Fig. 6， only吐1eareaAjof 白巴USPtape with d = 0.8 m m at c = 4000 % is larger than that of other three tapes. The larger th巴bendings廿ainamplitudeヲ白E smaller the fatigue crack growth area is. If the bending s仕ain amplitude is large， the fatigue crack growth area is small， resulting in a short fatigue life. In也ecase of USP wi白d=0.8 m m at c = 4000%， Ajis 1崎町幽other仕立E巴∞nditions，which coπesponds to the longer fatigue life. (2) F，αtigue crack growth 仰 Fat伊lecrack growth length As observed in Figs. 4 and 5，出巴fatigue企acturesurface shows a semi -elliptical or a quarter司ellipticalform. L巴tus discuss the fatigu巴crackgrowth length along也eflat surfaceas企om白e crack initiation point

F

，ιand that toward the center ac・The dependence ofas andac on the bending strain amplitudeんis shown in Fig. 7.InFig. 7， the data ofas紅巳plo出dby closed symbols and those of ac by open symbols， respectively. The regions， where也ey紅巳plo社ed，are shaded. As can be se巴n，th巴 larger the bending strain amplitude&，白血esmall巴rthe crack growth lengthas is.百1edependence ofac on 5a is slight. Values ofac are around 0.5 m m， a half of a thickness of the tape of 1 mm. Values ofas are larger白anthose ofac since也ebending 申込齢Hl!

，

←一一一月一一一、， R 噌ぺ X - 喝 d毛色ゐ.

"

s

-， F

'

"

h 抵手，、芝副会，

"

-

a

-f斗5llCfJ悶 lSf'世O.8X4制lO% I吾roO.SX2母00"1" むSP

o

1.2 x 2尋唱曲号も A判官eI¥'cd A 込話阜 4 J P

、

.

一場時 j } d 一 J e 一 e ，斗 { 一ら﹄ b -‘ d 一 3 M d M M M 一時ゆ緯 n

s

+ 句組九 2 廿 e s

一

u - o n i ? ﹂ u 一 n 自弘司 E 一 E 4 1 句，ゐ

。

_。

Fig. 5 SEM photographs of fracture surface ofUSP tape with d = Fig. 6 Relationship between fatigue crack grow出 area and bending s甘ainamplitude. 0.8 m m and c = 4000% for5a = 1.64% and J

ザ=

6788・(め whole fracture surface， (b) crack initiation p副 (1)ラ (c) middle pa此ofunstable企acture(2)ヲand(d) final unstable fracture p紅t(3).

(6)

88 愛知工業大学総合技術研究所研究報告，第17号， 2015年 " 思伽 . 旬、ι 2 ピ

_.

1.5 e

"

暢

'

"

' ・陶

"

-

4 包 ' h

:

由

5

=

::.t

.

-

_ー

"

，

.

_。

帥 r As...rlピ明日付 ". : Ak>ll!!surfacc ~ XSPt/J仏書X4韓日韓斗むデー." 〆.ti5P，p1.2 x 200唖?品

. .

'

血

4

・

H

・ .

・

.

〆ム USP同ぷx200(1'沌 1 2 3 4 B叩 dingstrai司縄問plitude 込 l~ン;，) 5 Fig. 7 Dependence of fatigu巴cracklen尉lSas aIld ac on bending 5仕ainamplitude:as is shown by closed symbolsaIld ac by open s戸nbols と 2 暗.6 .: 0 . 5 ι

E

さ

/

'

As-reccind 己主制 lIJNf~ / 苦言F

=

、

号、・

骨.

.

.~ ~

/

"

主そ

y O o， / 国ミ'{ / ，，， -歪色.ll.2，; * / " /' # 戸 ! ヲ . - . ，_ '" / 忌吾 /

，0'ジ

;J

，J - w F外相酬明，久/九

y

0.1 川刀;;0'

o

/

l

.

，

.

!

)

;

_

'

き

。

:

ィトァ

ι--".J九〉 1 町一一一」 D I 2 3 4 5 Bl.'制Ii鈴Z叫悶初amplirudeられぬ Fig. 8 Dependence of average fatigue crack growth rates a/Nfand

a

!

^

ケ

onbending strain arnplitud巴5afor as・receivedtape aIld USP tap巴w1血d=0.8 mm aIld c土 4000%:仏/N;is

shown by closed symbols anda!J'

ゆ

byopen sY1巾ols.

S仕ainis largest on吐1eflat surface of出巴tapeand is smaII in the central part ofthe tape. (iii)Average fatigue crack growth rate 百四 dep巴ndenceof average fatigu巴 crackgrowth rat巴tiIl 合ac旬realong the flat surfacea

〆

め

a

I

l

d白attoward the

∞

ntera

〆

め

onthe bending s廿ainarnpIi旬de5a for as明recei、吋tape

a

I

l

d the tape USP-仕切tedwithd = 0.8mm at c = 4000% is shown in Fig. 8.InFig.8， the data of

a

/

11ケareplo批dby closed symbols and

出oseof

a

!め

byopen symbolsラrespectively.Asωn be seen，吐1e

larger the bending strain ampliωde， the larger the crack grow也 rates are.The average fatigue crack growth rates for the as圃receivedtape紅'ehigher th

a

I

l

those for the USP tape. This

means the fact that也e average fatigue crack growth rates decreas巴bythe USP trea仕nent，resulting in longer fatigu巴Iife.

3.2.31;

ポ

uenceof hardness on

メ

It伊 el件 Vick巳rshardness of the surface for four kinds of tapes was measured by a load of 49 N.百lerelationships between the numb巴rof cycle to failure時andVickers hardnessHV for various values of bending stain arnpIitudeらar古shownin Fig. 9.In Fig. 9， the data are plo恥dby several symbols andせlerelation for each らisconnected by也es凶ightline. As can be seen in Fig. 9ラ也巴 fatigueIife

め

ofTiNi SMA 仕切ted by USP increases in proportion to Vickers hardnessHVas s鉱n巴asnormal metaJs. The influence ofHV on

ケ

^

isslight in也ecase of large5a・百1e relationships shown on th巴semi-logarithmicgraph are almost exp民ss巴dby straight lines for eachゐ Therelationship ther巴fore can be expressed by the foIIowing equation. log

め

=loga+bHV (3) where a andb町郎副めinHV= 0 and也eslope of the 10叫 -H V curve， respectively. The vaJue ofb denotes the sensibility of influence ofHV on the λ

T

;

The a denotes也evalue of

ベ

f

unaffected by HV. The fatigue Ii長Nfc組 beexpr悶巴dby an expone凶iaJfunction ofHV企omEq. (3) as foIIows:

め

=a exp(h110・bH

η

(4) The coefficientsa andb can be expressed by白elinear functions ofιas foIIows α=aj ら+σ2 b=bj5a+b2 (5) where a j and b j represent the slopes of thea -5a C町veand bーら curve， anda2加db2 repres巴nta組db in5a = 0， respectively. The values ofaj， a2ヲbjandb2 ar官・86.5，32.6，・0.0988and 0.00738ラ respectively. The results calculated企omEqs. (4) and (5) are shown by solid lines in Fig. 9.百leoveraJl inclinations are weII-approximated by the solid Iines. 3

ユ

4Surfiαceγoughness Itis known白atsurface roughness of也ematerial affects也巴 fatigue life. If theSl汀faceis rough， the fatigue life is short16).As observed in Fig.2， the fatigueIife of the poIished tape is longer than the as，・receivedtape. Inorder to investigat巴theinfluence of surface roughness on也efatigue life， surface roughness was measured. Arithmetic mean roughness Ra for five kinds of tapes is shown in Fig. 10. The value ofRa for吐1巴as-receivedtape is 0.65

(7)

as follows. The tap岱 weresu吋ectedto high work hardening during吐1erolling pro

∞

ss to produce them and therefore high residual stresses were already induced in the material.百1巴detai! ofthe residual stress is也efuture work.

o

l.2X2骨{紬呪，

o

仏書X4U告書 % 帥.SX2

制

0%. '/8.=1% Aト町'cdycd ，/ 0 ピ ~

• .

T

ろお2~.~ ~-aに_../"怠ら寸% I-I___"""'-<> tapes showing super巳lasticity

USP-treated， in which both flat surfaces were shot-peened from two opposite directions， and the influence of shot media diameter and coverage on the fatigue life of altemating-plane bending was investigated. The results obtained are summarized as follows. W巴re S乱

ι

屯 4. Conclusions TiNi Th巴 -Calculutcd 10'、ふ」 Z納 311抄 4樹) :;由。 ¥ickε時hardnc日1fI' 一、 a " I

"

-

_._._._. Z 芸1がシ

"

相続言語句，.

"

‘ ・・_;₎₁₀. き

J

Z

2 2 -r. 品00 Fig. 9 Relationship between number of cycles to failure and Vickers hardness for various strain ampliωdes&a・ The larger the bending strain ampliωd巴， the shorter the fatigue life is. The fatigue life can be expressed by a power function of strain amplitude. The fatigue of USP-treated tapes is longer也an也atof也eas-received tape. The fatigue lives of both tapes USP幽treat巴dwith different shot media diameters at the same coverage ar巴almostthe same. The fatigue life ofthe tape USP-treated with high coverage for the same shot media diameter is longer than也atwith

iii

r ' a

γ よ 1 1 ! i l t i 1.5

5

ぞ :::%1.1 俵羽詰 z

-r

.

11.'1 .

=

E Z昏.6 2 u 言。.3~

-=

車問

.

low coverage. The fatigue crack nucleates at the central part of也eflat surface of血etape in the case of也巴as-receivedtape. The fatigue crack nucleates at the comer near the flat surface of the tape in出ecase of the USP-treated tape. The average fatigue crack growth rate decr巴asesby the USP tr問団巴nt. The hardness on the surface of仕1etape increases by the USP仕eatrnen.The ft atigu巴lifeincreases in proportion to the hardness on the surface of也etape for each bending strain amplitude. The fatigu巴lifecan be expressed by an exponential負担lctionofVickers hardness. The surface roughness increases slightly by the USP treatrnen The f.t atigue life of polish巴dand USP-treated tapes is longer than that of the as-received tape. Inpractical applications of SMAs， the fatigue life of SMA elements increおおifUSPis treat巴dnot only on the surface at the maxinmm s佐巴sspoint but also on the surface in the region near the maximum stress point. Acknowledgment

The experimental work for this study was carried out with the assistanc巴ofstudents in AichiInstitute of Tec加1Ology，to whom the authors wish to expr巴sstheir gratitude. The authors 2. 3. 4 5. Fig. 10Ari社uneticmean surface roughness for five kinds of tapes. p，必hhed 戸n.百1evalue ofRa= 0.2卯nfor the polished tape is about one-thirdぉlargeas也eas-received tape. As observed in Fig.

乙

the fatigue life of the polished tape is longer thanせ1atofせ1e as-received tape. Thereforeラif吐1巴surfaceis smoo白，the fatigue life is long・Thevalues of

R

a

for three kinds ofUSP tapes紅巳0.83 問1-0.88Jllll， which蹴 slightly larger th粗削 ofthe as-received tape. As observed in Fig. 2ラthefatigue life of USP tapes is longer th加 thatof the俗画receivedtape. Compressive residual s仕己ssesare induced in the surface layer of the material by USP 8).Although the surface roughness increases slightly by USP， the influence of USP-induced surface roughness on也E fatigue liたisslight. The reason why the USP・inducedsurface roughness has a small influenc巴on血εfatiguelife must be due to compressive 1巴sidualstresses induc巴dby USP in the sぽfac巴layer

of the tape. The values of residual str巴ssmeasured were among ・80MPa叫ー160MPa in the 岬 onofdep也企omthe surface to 200μm for both the as-re

∞

ived tape釦dUSP tapes. The

differenc巴ofthe residual str巴sses恒nong也eas-received tape and

USP tapes was not clearly detected. This reason may be surmised

..¥....re"...hcd 唖

(8)

90 愛知工業大学総合技術研究所研究報告，第17号， 2015年 also wish to巴xtendthanks toth巴administratorsof Scientific Research (C)(General)加Grants-in-Aidfor Scientific Research by仕leJapan Society for Promotion of Science for fmancial support. REFERENCES 1) K. Otsuka and C.M. Wayman， eds.， Shape Memory MaterialsラCambridgeUniversity Pr巳ss，Cambridge， pp. 1・49，1998. 2) T. Sakuma， U. Iwata組dY. Kimur，aCyclic Behavior and Fatigue Life of TiNiCu Shape Memory Alloyヲ FATIGUE '96，1， pp. 173-178ラ1996

3) T. Sawaguchi， G. Kaustrater， A. Yawny， M. Wagner and G. EggelerラCrack也itiationand propagation in 50.9 at pct Ni-Ti pseudoelastic shape-memory wires in bending-rotation fatigue， Metallurgical and Materials Transactions A， 34A， pp. 2847-2860， 2003 4) M. Wagner， T. Sawaguchi， G. Kaustrater， D. Hoffken and G. Eggeler， Structural fatigu巴ofpseudoelastic NiTi shape memory wires， Materials Science and Engineering A， 378， pp. 105-109，2004. 5) R. Matsui， H.Tobushi， Y. Furuichi and H. Horikawaヲ Tensile Deformation and Rotating-Bending Fatigue Properties of a Highelastic Thin WireヲaSuperelastic Thin Wireヲanda Superelastic Thin Tube ofNiTi AlloysヲTrans ASME， J. Eng. Mater.Tech.ヲ126，pp. 384-391ヲ2004. 6) H. Tobushi， R. Matsui， K. Takeda and E.A. Pieczyska， Mechanica1 Properti巴sof Shape Memory Materia1s， Part 2. Fatigue properties of shape memory alloyラ Nova Sci巴ncePub.， New York， pp. 115-164，2013. 7) L. Wagner ed. Shot peening， Wiley-VCH， W出heirn，pp. 1・562，2003.

8) S.K. Cheong， D.S. Lee， J.H.Lee， M. Handa and Y.

Watanab巳ヲ Effect Of Ultrasonic Shot Peening on The

Fatigue Characteristics of Welded Sts304 for Rolling Stock， Proc. ICSP-I0， pp. 494・498ラ2008.

9) H. Tobushi， K. Mitsui， K. Takeda， K. Kitamura and Y. Y oshimi， Performance and Design of Precision-Cast Shape Memory Alloy Brain SpatulaラJ.Theoretical and Appl.Mech吋50-3，pp. 855・869，2012. 10) K. Taked，aH. TobushiラK.Miyamoto and E.A. Pieczyska， Superelastic Deformation of TiNi Shape Memory Alloy Subjected to Various Subloop Loadings， Mater.Trans.ヲ 53・l，pp.217・223ラ2012. 11) H. Tobushi， T. Nakahara， Y. Shimeno and T. Hashirnotoラ Low-Cycle Fatigue of TiNi Shape Memory Alloy and Formulation ofFatigue LifeラTrans.ASMEラJ.Eng. Mater. Tech.， 122， pp. 186-191，2000 12) E.A. Pieczysk，aS. Gadaj， W.K. Nowacki， K. Hoshio， Y. M泳inoand H. Tobushi， Characteristics of energy storag巴 and dissipation in TiNi shape memory a1loy， Sci. Tech. Advanced Mater.， 6， pp. 889司894，2005.

13) R. Matsui， Y. Makino， H. Tobushi， Y. Furuichi and F. YoshidaヲInfluenceof Strain Ratio on Bending Fatigue Life and Fatigue Crack Growth in TiNi Shape贋Memory Alloy Thin Wires， Mater.Trans.， 47・3，pp.759・765，2006. 14) K. Tanaka， T. Hayashi， Y. Ito and H. TobushiヲAnalysisof th巴rmomechanical behavior of shape memory alloysラ Mech. Mater.， 13， pp. 207・215，1992 15) K. TakedaラK.Mitsui， H. TobushiヲN.Levintant-Zayonts and S. Kucharski， Influence ofNitrogen IonImplantation on Deformation and Fatigue Properties of TiNi Shape Memory Alloy Wire， Arch. Mech. 65-5， pp. 391-405ラ 2013 16) J.E. Shigley and C.R. Mischke， 5由巴d.，Mecharせcal Engine巴ring Design， McGraw-Hill， New Y ork， pp. 282開283，1989.