NiTiCu shape memory alloy tested on a “soft” torsion machine

NiTiCu shape memory alloy tested on a “soft” torsion machine

Scripta METALLURGICA NiTiCu SHAPE Vol. 22, pp. 1 0 2 3 - 1 0 2 8 , 1988 P r i n t e d in the U.S.A. MEMORY ALLOY TESTED ON A "SOFT" P e r g a...

314KB Sizes 0 Downloads 111 Views

Scripta

METALLURGICA

NiTiCu

SHAPE

Vol. 22, pp. 1 0 2 3 - 1 0 2 8 , 1988 P r i n t e d in the U.S.A.

MEMORY

ALLOY

TESTED

ON A "SOFT"

P e r g a m o n Press plc All r i g h t s r e s e r v e d

TORSION

MACHINE

K. GOUBAA, V. ORLIONNET, M. MASSE, G. BOUQUET Laboratoire de M~tallurgie Structurale, E.N.S.C.P.- C.E.C.M. E.N.S.C.P., II rue Pierre et Marie Curie, 75231 Paris Cedex 05 C.E.C.M., 15 rue Georges Urbain, 94400 Vitry-sur-Seine ( R e c e i v e d F e b r u a r y 22, 1988) ( R e v i s e d M a y 4, 1988) Introduction A particular apparatus was designed and perfected in our laboratory to study the thermomechanical behavior of a NiTiCu alloy aimed at orthodontic applications (])(2)(3). For such an alloy, the transition temperatures As and Af must be placed lower than the temperature of the mouth (approximately 35°C) while Ms and Mf must be above 25oC to allow metal forming processes at room temperature, i.e. in the martensitic phase (4). The thermo-mechanical characterization of shape memory alloys would give information, on the one hand, about the shift of the transition temperatures under the effect of a static stress and, on the other hand, about the dependence of the dynamic stress on the deformation of the material. In this aim a "soft" torsion apparatus was built to perform tests, the metallurgical part of which is presented here. Experimental Device The scheme of the "soft" machine is reported in Fig. I. The terminology "soft" refers to the fact that on such a machine, under an applied stress or under the effect of the temperature, the specimens undergo a free deformation. In our case a torsion deformation mode is used and applied to samples (l to 2 ann in diameter and 20 to 30 mm long) with the help of a driving-wheel system. The loading consists of 3g steel balls distributed one by one and falling in bags positioned in such a way that either a loading or unloading stage is possible. In these conditions, torque increments by steps of l,l N.mm are available. The furnace consists of a heating resistance immersed in a bath of silicon oil. This bath can be also cooled by evaporated nitrogen flowing in a copper pipe encircling the oil container. This assembly permits investigations in a temperature range spreading from -40 to 200°C. Desired slopes of heating or of cooling rates are supplied by a temperature programmer. With such a device we can draw curves such as a/l=f(T) at constant stress ( a : torsion angle, I: length of the sample) or a/l=f(M) at constant temperature, M being the torque applied to the specimen. Experimental Results Evolution of the deformation a/l as a function of the temperature T A shape memory alloy of chemical composition Ni 50,5%, Ti 44,5%, C u S Z ( w e i g h t p e r c e n t a ges ) was prepared in a high frequency furnace under argon-atmosphere. With respect to the transition temperatures of equiatomic NiTi alloy, the Cu alloying element shifts structural transitions towards low temperatures (5). Samples of this alloy were tested on the "soft" torsion machine to obtain a/l=f(T) curves, which are reported in Fig. 2. The specimens were prior annealed in the austenitic phase (8000C-30 min.) and then quenched in water. Once in position on the torsion machine, the sample was loaded at low temperature (-40°C). The desired load being reached, after a mechanical stabilization of the sample, the deformation changes were followed as a function of the temperature either on heating or on cooling. On heating, in the low temperature part of the curves a/l=f(T) in Fig. 2a, the deformation increases slowly and then accelerates until it reaches a steady rate associated with an easy deformation stage. Above 50°C, the sample strain decreases while the temperature is still increasing. This opposite evolution of the deformation must be related to the modulus change occurring when the material undergoes the phase transformation from martensitic to austenitic form. As indicated on the curve in Fig. 2a, this evolution enables us to determine the transi-

nn~6-9748188

1025 $3.00

+ .00

1024

SHAPE

MEMORY

Vol.

22,

No.

7

tlon temperature A~ (austenite start transformation under stress a). The maximum appearing in the curve ~/l=f(T) is probably due to two competitive phenomena acting on both sides of this maximum. Between 70 and |00°C, the deformation decrease slows down, indicating the end of the martensite to austenite transformation. Consequently, above 50°C, the structure of the material comprises two phases: austenite and martensite. On the cooling part of the deformation-temperature cycle, a modulus evolution arises, opposite of that occurring on heating and giving rise to a deformation increase as the martensite appears at the temperature My (Fig. 2a). We note that this first e/l=f(T) cycle is not closed, indicating a final deformation state greater than the initial one. This phenomenon is discussed later. A second ~/l=f(T) cycle, obtained under the same torsion stress as in the previous case, is reported in Fig. 2b. Firstly, we notice that the cycle is perfectly closed and, moreover, the curve part drawn on cooling can be superimposed upon the corresponding one obtained in Fig. 2a. This observation agrees with a structural stabilization resulting from the thermal treatments of the first cycle. Secondly, which may be the main feature of this experiment, we observe the arising of anomalies consisting of "shoulders" on the curve, both on heating and on cooling. On heating the "shoulder" occurs just before the martensite to austenite transformation is achieved, while on cooling the anomaly appears in the first steps of the reverse transformation. This result accounts probably for an additional reversible structural change (Fig. 3). Lastly, a third ~/l=f(T) cycle is carried on in zero loading conditions, therefore the material is tested in a pre-strained state. The most interesting result is the deformation recovery observed on cooling. The effect is weak because of the lack of applied static stress and furthermore it disappears completely during the next temperature cycle (Fig. 2c). Effect of prior-annealing on the deformation-temperature cycles Changes in prior-annealing temperatures lead to modification in the deformation-temperature cycles as shown in Fig. 4. Indeed, we note that the height of the pseudo maximum, previously described, is modified as a function of the annealing temperature. This evolution is obviously consistent with a structural modification of the material but we must point out the lack of a well defined relation between maximum height and annealing temperature. In Fig. 4 the ~/l=f(T) curves are drawn relative to the same reference deformation level, while in Fig. 5 the same resuits are presented in such a way that the final deformation, in the austenitic phase, is invariant. In this case we can observe that the deformation rate, between As and Af, reaches approximately the same value, whatever the annealing temperature is. Between Ms and Mf a similar resuit is observed on cooling. Discussion On the curves e/l=f(T), the pseudo maximum observed on heating, at about 50°C, can be explained by the occurrence of two successive phenomena. The first one, leading to a deformation increase faster than a linear rate, corresponds to an easy deformation of the specimen above a critical temperature. This result was related to the rubber-like behavior of the material, generally consisting of reorientation of martensitic variants in order to have their most favourable orientation with respect to the applied stress (6). At constant temperature but at increasing load, this phenomenon starts when a sufficient stress is reached. In our experiments, the load being constant, the re-organization of variants will be therefore dependent on the increasing temperature. Thus the effect of the temperature will censist of decreasing the critical stress required for the reorientation of the various martensitic variants. Consequently, on ~/l=f(T) curves, a specific temperature, depending on the variant misorientation, will have to be reached in order to allow the reorientation of these martensitic variants. Above the critical temperature, the deformation is easier, leading to an acceleration of the process. Above 50°C, the second phenomenon, corresponding to a deformation decrease, was associated with the occurrence of the austenitic phase, the shear modulus of which is higher than that of the martensitic phase. Consequently, the pseudo maximum observed on ~/l=f(T) curves can be explained in terms of two successive effects: rubber-like behavior and modulus change. The decreasing deformation associated with this last effect allows the determination of the transition temperature A~. Conversely, at the temperature M~, an important deformation increase takes place associated with an opposite modulus effect, on cooling. When annealing temperatures are changed, the result which can be expected is a modification of the initial structural state and particularly the repartition of the martensitic variants. More precisely, in our experiments, changing the annealing temperature can induce, on cooling, a distribution of martensitic variants more or less divergent from random. Thus, the

Vol.

22,

No.

7

SHAPE

MEMORY

1025

material will be partly composed of misoriented martensitic variants with respect to the stresses developed during a torsion test. Consequently, before reaching A~, misoriented variants will have to reorient under the combined effects of stress and temperature, to satisfy the shear stress components. The height of the maximum associated with the rubber-like effect, will be higher as the rate of misoriented variants is important, as shown in Fig. 6. A second characteristic of the first ~/l=f(T) cycle is the non-closed shape which was attributed to a structural stabilization. Indeed, during the cooling stage of the cycle, the specimen undergoes, for the first time, an austenite to martensite transformation under stress. In these conditions, the distribution of the martensitic variants is somewhat different from the initial one resulting from quenching followed by loading in the martensitic phase. Thus, after the first temperature cycle under loading, the structure is as "educated" with respect to temperature and stress parameters, so that the next cycles must be reproducible. This is what we observe on the curve of Fig. 2b: the cycle is closed and others cycles drawn in the same experimental conditions can be superimposed. Concerning the third temperatUre-deformation cycle (Fig. 2c), in zero loading conditions, the main feature is the little deformation appearing on cooling and which was attributed to the reversible shape memory effect. This phenomenon is not observed on further cycles. An additional result to explain is the "shoulders" observed on the curve of Fig. 2b. It was pointed out that, with regard to the positions of these anomalies relative to the transition temperatures, a reversible structural phenomenon could be implied: on cooling the anomaly occurs before the austenite to martensite transformation, while on heating, the phenomenon is observed in the ending stage of the opposite phase transformation. It is well known that in NiTi alloys, pre-martensitic transformations have been reported (7)(8). Such phenomena could explain the anomaly observed for our alloy in spite of a somewhat different chemical composition due to the Cu alloying element. It must be said that the anomalies previously described are not always detected and two assumptions can be put forward to explain this result. The first one is the possibility of a lack of sensitivity of the torsion machine in detecting phenomena that are occasionally weak. The second one would relate the non-observed anomaly to an effect resulting from specific quenching conditions or to the nature of the annealing treatment before quenching. These assumptions have to be clarified by others experimental procedures. Precisely, internal friction experiments are in progress in order to throw light on this problem. Conclusion The results presented here do not concern the most favourable alloy aimed at dental applications, however, the metallurgical behavior of this material is interesting. Our study has shown that a "soft" torsion machine is a valuable tool in studying the mechanical evolution of shape memory alloys. Indeed, with such an apparatus, under a constant load, the curves, giving the deformation of the material as a function of the temperature, exhibit two main characteristics. The first one is a maximum in the deformation evolution, at about A~ temperature, which was related to the occurrence of two successive phenomena. Below A~, in martensitic phase, after a first strain corresponding to the loading of the specimen, the deformation increases, under a constant stress, as a function of the temperature. This evolution was attributed to a rubberlike effect acting progressively under the influence of the temperature. The reorientation of martensite variants, implied in the phenomenon, leads to an additional deformation in the direction of the applied stress. Above A Os, the austenitic phase appears and the resulting increase of shear modulus involves a decrease of deformation, explaining the observed maximum in this deformation evolution. Moreover, it was shown that the prior-annealing temperature affects the height of this maximum. The prior-annealing temperature was assumed to modify the orientation distribution of martensite variants and the maximum height was associated with the rate at which these variants reorientated with respect to the stress developed in the torsion test. The second characteristic is the appearance of "shoulders" on deformation-temperature curves. These anomalies are, respectively, detected on the heating and cooling parts of the deformation as a function of temperature cycles. The "shoulders" occur, on heating, during the ending stage of the martensite to austenite transformation, and on cooling, during the first steps of the reverse transformation. Consequently, this phenomenon w a s a t t r i b u t e d to an a d ditional reverse structural transformation, perhaps as a p r e - m a r t e n s i t i c stage. Internal friction experiments are in p r o g r e s s to c l a r i f y this b e h a v i o r .

1026

SHAPE

MEMORY

Vol.

22,

References |. 2. 3. 4. 5. 6. 7. 8.

G.F. Andreasen, P.R. Brady Angle Orthod. 42 (1972) 172. G.F. Andreasen, R.D. Barrett Am. J. Orthod. 63 (1973) 462. G.F. Andreasen, Am. J. Orthod. 78 (1980) 528. Shape Memory Effects in Alloys J. Perkins Ed. Plenum Press, New-York (1975). K. Melton, O. Mercier Acta Met. 29 (198l) 393. L. Delaey, R.V. Krisnan, H. "Tas, H. Warlimont J. of Mat. Sci. 9 (1974) 1521. P. Moine, G. Michal, R. Sinclair Acta Met. 30 (1982) 109. Li Xuemin, T.Y. Hsu Mater. Sci. Eng. 91 (1987) 189.

D

FIG. 1 Scheme of the torsion machine: I- counter-weight 2- suspension wire 3,3'- driving-wheels system transmitting torsion stresses to sample 4- receiving balls bags permitting loading of sample 5- steel balls distributor 6- sample 7- thermal insulator 8- copper pipe for evaporated nitrogen flow 9- heating resistance I0- silicon oil.

No.

7

Vol.

22, No.

ct/1

7

SHAPE MEMORY

1027

deg/mm

FIG. 2

.-T,~

4

~ "

~

Deformation-temperature cycles a static torque of 132 N.mm:

\

under

0J

" \ - ~. : l ~



~O

~3 G o

2a first cycle exhibiting a maximum and a non-closed shape --.--

2b second

cycle perfectly

closed

~2

........ 2c third cycle in zero loading conditions: reversible shape memory effect on cooling.

O

...................... q .....

~,~.~

M~ S I

i

i

|

i

i

0

~/I

I

i

i

i

|

50

i

T°C

lO0

deg/mm ..........."~

FIG. 3

4

Set

~%

oJ

"~o0 =O ~ =3

~'~

~ ~

i

\<~.~

~

f

l

~/I

"shoulder"

~--

i

0

second

cycles, i!

~\

i

successive

un-

der a static torque of 132 N.mm, • showlng the appearance of shoulders anomalies on heating and on cooling.

~ ~ .

on cooling

of

I

,

i!

on heating

~

i

I

i

i

50

|

i

TOC

]00

deg/mm FIG.4 Deformation-temperatures cycles drawn relatively to a same initial deformation level: effect of priorannealing temperature on the height of the deformation maximum:

~5 nO

~4 0 "'~ 09

500°C ....... 600°C 650°C

3 .

o 4-J

--.--

2 ~,~~,=-::::'-:-,

|

*

0

*

i

i

50

i

,

i

i

......

-

100

-

*

TOC

-

-

-

_

700°C 700°C

1028

SHAPE

=/I

MEMORY

Vol.

deg/mm

~5

1.1

r~

~

QI

~ .~

~

/ /

0

~

O

~ ~ ? ,~



-

~l ~!



o





l

i

\\\X \\\%

i

i

i

® "~

i

..4

h

:3

=

.4

i

I0O

50

i

T°C

FIG.5 Deformation-temperature c y c l e s as in l a t i v e l y to a same final d e f o r m a t i o n r e f e r e n c e . S i m i l a r d e f o r m a t i o n rates h e a t i n g and on c o o l i n g , w h a t e v e r the t e m p e r a t u r e is.

Fig. 4 d r a w n res t a t e t a k e n as a are o b s e r v e d on prior-annealing

o

L

Temperature FIG.6 S c h e m e of the m a x i m u m d e f o r m a t i o n h e i g h t as a f u n c t i o n s t r u c t u r a l s t a t e r e s u l t i n g from q u e n c h i n g |["

[ rate pect

of

the

of m a r t e n s i t i c v a r i a n t s w e l l - o r i e n t e d w i t h resto s t r e s s e s and t e m p e r a t u r e g r a d i e n t p a r a m e t e r s

rate of m a r t e n s i t i c v a r i a n t s to the same p a r a m e t e r s .

misoriented

with

respect

22,

No.

7