INTERNAL FRICTION MEASUREMENTS RELATED TO THE TWO WAY MEMORY EFFECT IN Cu-Zn-Al ALLOY EXHIBITING THERMOELASTIC MARTENSITIC TRANSFORMATION

(1)

HAL Id: jpa-00221029

https://hal.archives-ouvertes.fr/jpa-00221029

Submitted on 1 Jan 1981

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

INTERNAL FRICTION MEASUREMENTS RELATED

TO THE TWO WAY MEMORY EFFECT IN Cu-Zn-Al

ALLOY EXHIBITING THERMOELASTIC

MARTENSITIC TRANSFORMATION

M. Morin, G. Guenin, P. Gobin

To cite this version:

(2)

JOURNAL DE PHYSIQUE

CoZZoque C5, suppl6ment au nolO, Tome 42, octobre 1981 _page_C5-1013

INTERNAL FRICTION MEASUREMENTS RELATED TO THE TWO WAY MEMORY EFFECT

IN Cu-Zn-A1 ALLOY EXHIBITING THERMOELASTIC MARTENSITIC TRANSFORMATION

M. Morin, G. Guenin and P.F. Gobin

Groupe d'Etudes de Me'taZZurgie Physique e t de Physique des Matdriaux

-

E.R.A. 463, I n s t i t u t NatimzaZ des Sciences AppZiquges, B6t. 502,

69621 ViZZeurbanne Cedex, France

Abstract.- The alloys vitL T~lernioelastic Martensicic Transfornation, like the Cu-Zn--el, exhibit, a£ ter a thermomechanical treatment named "training", The Two Way Shape Memory Effect. We have compared internal friction measurements versus amplitude for a "trained" and a "virgin" sample. Tie obtained results are coherent with the existence in the "trained" saruple of particular sets of dislocations. These dislocations generates orientated internal stresses, which favour particular martensite variants during the direct transformation.

1. Introduction.- The anelastic characteristics of alloys exhibiting Phermoelastic Iv!artensitic Transformation are now relatively well known:a large damping is oiserved in the martensitic temperature range and a very high peak appears during the transformation itself.Some new features which concern especially isothermal nleasurements in the transformation range and the effect of mechanical cycling have been published elsewhere Ly the authors (1) (2).

The present work relates the changes in tiie anelastic behaviour due to the "training" of the.sample ; this "trainins" being a special treatment which induces a two way memory effect in the sample (3)(4)(5).

2. Apparatus, sample preparation.- Internal friction measurements have been perfor- med on a torsional pendulum at approximately one cycle per second by determination of logarithmic decrement-6 of free oscillations. The sample preparation has been des- cribed elsewere (1). The atomic composition of the sample was Cu, 13.6 at % Zn,

l i at Z A1 (I~S = 280 K)

.

3. Results.- The Two Way Memory Effect treatment : tile "training" process was the following : In the high temperature phase

(B

phase) the sample was submitted to a small torsion stress. T'nis stress was well below the yield stress of the

6

phase and not sufficient to induce the martensite transformation at the temperature involved (tne corresponding maximum elastic strain is E ) . When the temperature is lowered the sample exhibits a large strain in the direction of the stress because the martensite variants which appear are the ones favoured by the stress. At low temperature (mar-

tensitic phase) a large strain Ef is therefore observed. This strain E~ is proportio- nal t o € , at least for small E (figures 1 and 2).Vhen tiie temperature is raised, the strain comes back; this fact is due to the shape memory effect (martensite to

B

phase transformation).

(3)

JOURNAL DE PHYSIQUE

Fig. 1 : Torsion strain induced by static Pig. 2 : Ultimate strain inducea &

f torsion stress during thermal transformation as a function of the initial elastlc cycles. Several stresses in the

B

phase are strain &

e' used giving several elastic strain E

.

For the "training" operation the sample undergoes several such thermal cycles, with the small torsion stress applied on. The figure 3 shows the curves

E = f(T) obtained for several successive thermal cycles.

Pig. 3 : The "training" thermal cycles C1 to 14) and the cycle 15 whithout applied stress showing the "two way memory effect".

Some observations are to be done :

-- The beginningof the first transformation at TIS1 is raised by the torsion stress application. This phenomenon is well know (6).

(4)

observed with the number of thermal cycles. This phenomenmsaturates after a few cycles.

-

When this "trained" sample is now cooled again, without any applied stress a spontaneous strain is observed during the martensitic transformation in the same direction than before, but this strain is smaller. This strain disappears during reheating by shape memory effect. The sample exhibits now the "two way shape memory effect". Finally two importante facts are also to be noticed :

-

i : the temperature of the beginning of the transformation MSI5 is rai-

sed in relation to the one of the virgin sample M SO'

-

ii: the sample free of stresses in the

6

phase keeps a small rernanent strain E

.

r This behaviour seems to show that internal stresses induced during the training are superimposed to the applied static stress. Besides, when the static stress is removed (cycle nomber 15) these internal stresses substitute themselves to the applied one, and partially orientate martensite variants, giving spontaneous strains, during the cooling ("two way memory effect").

The remanent strain observed E seems to show that these internal stresses are due to dislocations. This fact is confirmed by an electron microscopy study of the f3 phase of a "trained" sample : by contrast with "virgin" sample where the dislocation density is very low ( 7 ) , the "trained" sample exhibits in the 6 phase many pa- rallel dislocations put as in flexion boundaries (figure4).

Fig. 4 : Dislocation struc- ture in 6 phase for a "trained" sample

These dislocations were not induced by classical cold work in the 6 phase because the "training" static stress was always very much smaller than the usual yield stress It is to be noticed than when the temperature decreases, this stress turns to be sufficient to induce some particular martensite variants giving rise to large strain of the sample (fig.3). These special sets of dislocations are induced during the nucleation or more probably during the growth of these particular martensite variants.

(5)

C5-1016 JOURNAL DE PHYSIQUE

cycled but without applied static stress is probably more in the dislocation ordering than in the dislocation density. If the dislocations induced by the successive transformations are randonly orientated (following the random repartition of the martensite variants in an unstressed cycled sample), the mean strain will be nul. However, if the transformation induced dislocations are orientated during the "training", the strain of the sample in the f3 phase will be significant (E ) . As we noticed before, the internal stresses produced by these orientated dislocations are probably responsi- ble for these-reverse memory effect" (memory effect during the f3 to martensite transformation).

V l R O l N SAMPLE \RAINEO" SAMPLE

I I I 1 I I I I I I

0 1 1 3 4 5 0 1 2 3 4

A after

-

10 measurin cycles at E max"5.10

5

B after= 100 measuring cycles at E max = 5.

lo4

C after=500 measuring cycles at E max

-

5.10-4

Fig. 5 : Curves 6 = f (E) in isothermal conditions for the "virgin" sample (a,b,c) and the "trained" one

(6)

- Internal friction measurements in isothermal conditions : Three temperature range have been chosen :

- Near the beginning of the transformation (T = M )

-

In the temperature transformation range (Mf < T < Ms)

- In the martensite temperature range (T < Ms)

The variations of internal frcition

6

with the maximum strain E have been measured in free decay conditions. For every temperature the curves 6 = f(~) have been drawn after approximately 10, 100 and 500 mechanical cycles of the sample at

-4

maximum strain amplitude (E = 5 x 10 ) . The figare (5) shows the results obtained m

with a "virgin" sample and with a "trained" sample. A noticeable difference is only observed at the beginning of the transformation where the "trainingu induces a "peal?' of 6 at low strain amplitude (fig. 5a1). This effect seems to increase with the mechanical cycling of the sample at the maximum strain amplitude. This peak could be due to an interaction between some orientated dislocations induced by the training and the first martensite plastes. The decrease of

6

with the number of mechanical cycles have already been observed (2) and is probably due to the progressive pinning of interfaces which move under the action of the measuring stress.

-

Internal friction measurement at slowly decreasing temperature. The curves

6

= f (E) drawn when the temperature is very slowlyf: decreasing in the transformation temperature range (Mf < T < M ) are modified in relation to the same curves drawn in isothermal conditions. For the "virgin" sample (fig. 6a) an increase of 6

-5

is observed in the whole range of the strain amplitude (0 < E < 6x10 ) , this increase being much larger at low E. For: the "trained" sample (fig. 6b) the effect is very small (if any) even at low E. This new component of internal friction called I.F.3 in ref.(2) can be obtained by difference between the curves&= f(E) drawn at constant temperature (curve 1, fig. 6) and the ones drawn with decreasing temperature (curve

2, fig.6).

It has been observed that this component IF3 rises at small val~es of E

in the "virgin" sample (fig.6a). If this internal friction is interpreted in terms of Delorme and Gobin's model (8) this rise significates that below a threshold strain amplitude the quantity of orientated martensite during a mechanical cycle is a constant and independant of strain amplitude. Above this critical strain amplitude the quantity of orientated martensite during a mechanical cycle would be proportionnal to the strain amplitude.

In the case of the "trained" sample it can be seen that the componant IF 3 due to the temperature decrease p r a c t i c a l l y v a n i s t r e s . T h i s could be due to the fact that the internal static stresses generated by training" are higher that the periodic measuring stress. The martensite, therefore is only orientated by the internal static stresses. It must be noticed here through premiminary experiments that thermally

---

(7)

JOURNAL DE PHYSIQUE 60

-

_V_{I R G l N SAMPLE} ₆₀

-

'TRAINED" SAMPLE Mf<T<Ms 5 0 - 0.05 K h n (2) \ 20- \ 10

-

- - -

10 I I I I I I 0 '2 .3 . 4 .5 4 0 ,1 .2 .3 . 4 .5 .6 * E A O - 4 EAO-.

Fig.6 : curves 6 = f(~) in isothermal conditions (curves 1) and for slowly decreasing temperatures (curves 2). The I.F. difference between the curves 1 and 2 (IF3 in ref.

(I) (2)) has been represented by interrupped lines. a = "virgin samplef'; bztrained sample"

_---

cycled samples without any stress also exhibit a drecrease of I.F. 3'

5. Conclusion.- The alloys with thermoelastic martensitic transformation can exhibit the "two wag memory effect" after particular thermomechanical treatment. This effect can be induced by "training" (successive transformations under static stress). Some effects of this "training" on the anelastic properties have been measured. The obtained results are coherent with the existence in the "trained" sample of a particular set of dislocations whitch generates orientated internal stresses. These internal stresses favour particular martensite variants and prevents the orientation of the martensite by measuring stress during experiments at slowly decreasin temperature.

REFERENCES

1. M. MORIN, G. GUENIN and P.F.GOBIN, Internal friction and ultrasonic attenuation in solids, e.d., C.C.SMITH, Pergamon Press, July 1980, p. 275-280

2. M.MORIN, G. GUENIN, S. ETIENNE and P.F. GOBIN, Trans. J.I.M.,1981,~01.22~1,~.l 3. T.A. SCHROEDER and C.M. WAYMAN, Scripta Met., 1977, vol.11, p. 225-230

4. L. DELAEY and J.THIENEL, Shape Memory Effect in alloys, Plenum. Press New-York 1975, p.341

5. J. PERKINS, G.R. EDWARDS, C.R. SACH, J.M JOHNSON and R.R.ALLEN, Shape Memory Effect in Alloys, Plenum Press, New York, 1975, p. 273

6. L. DELAEY, R.U.KRISHNAN, H.TAS and H.WARLIMONT, J. of Mat. Science, 1974,vol. 9, p. 1521-1551

7. C. MAI, G. GUENIN, M. MORIN, F. LIVET and P.F.GOBIN, Mat. Sci. and Eng.,1980, vol. 45, p. 217-220