THE β↔2H MARTENSITIC TRANSFORMATION IN A Cu-Zn-Al ALLOY AS STUDIED BY SIMULTANEOUS ENTHALPY AND ACOUSTIC EMISSION MEASUREMENTS

(1)

HAL Id: jpa-00222217

https://hal.archives-ouvertes.fr/jpa-00222217

Submitted on 1 Jan 1982

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.

THE β ↔ 2H MARTENSITIC TRANSFORMATION IN A Cu-Zn-Al ALLOY AS STUDIED BY

SIMULTANEOUS ENTHALPY AND ACOUSTIC EMISSION MEASUREMENTS

A. Planes, J. Macqueron, M. Morin, G. Guénin, L. Delaey

To cite this version:

A. Planes, J. Macqueron, M. Morin, G. Guénin, L. Delaey. THE β ↔ 2H MARTENSITIC TRANS-

FORMATION IN A Cu-Zn-Al ALLOY AS STUDIED BY SIMULTANEOUS ENTHALPY AND

ACOUSTIC EMISSION MEASUREMENTS. Journal de Physique Colloques, 1982, 43 (C4), pp.C4-

615-C4-620. �10.1051/jphyscol:1982497�. �jpa-00222217�

(2)

THE e#2H MARTENSITIC TRANSFORMATION IN A C u - Zn- A l ALLOY AS STUDIED BY SIMULTANEOUS ENTHALPY AND ACOUSTIC EMISSION MEASUREMENTS

A. Planes*, J.L. Macqueron** , M. Morin** , G. Guenin** and L. Delaey***

*Fasultat de F-Csiea, Département de Termologia, U.B., Diagonal 645, Baroelona 28, Spain

**Institut National des Soienaes Appliquées de Lyon, Bâtiment 502, 69621 Villeurbanne, France

***Département Metaalkunde, Katholieke Universiteit Leuven, de Croylaan 2, B-3030 HeVerlee, Belgiwn

(Accepted 9 August 1982)

Abstract.-Simultaneous enthalpy and acoustic emission measurements were used to study the g-*2H martensitic transformation in a Cu-Zn-Al alloy. The results are compared with those obtained for the g-«-18R transformation in a Cu-Zn-Al alloy with different composition.The prominent results can be sumarised as follows: the transformation has a jerky growth character and signals obtained are not reproducible from one thermal cycle to the other. The thermoelastic character of this transformation is questionable.

Introduction. -It is well known that a large amplitude acoustic emission (A.E.) is generated during martensitic transformations (1-3). Several authors have used the A.E.

technique to study the kinetics of the thermoelastic transformation and also the pos- sible changes of internal structure of the high temperature phase induced by the transformation during thermal or mechanical cycling (4-6).

In the present work the acoustic emission technique has been used together with the calorimetric technique which allows to obtain simultaneously, for the same sample, the thermal power and the A.E. (7). The first results obtained by such a combination are related to the study of martensitic transformation of g Cu-Zn-Al alloys. The structure of the martensite phase in these alloys is composition depen- dent : for high aluminium concentration the martensite is y' with 2H structure; for low aluminium concentration the martensite is 6'with 9R or 18R structure (8).

What concerns g — 18R transformation, the most prominent effects measured by this method are (6):

a) Sharp thermal and acoustic peaks are superimposed on a background effect. These thermal and acoustic peaks are well correlated in temperature.

b) The apparence of both signals, A.E. and calorimetric, clearly evolute during thermal cycling. With increasing number of cycles the Ms temperature raises, the thermal and acoustic peaks decreases in height, the transformation spreads in temperature, and the total A.E. (continuous and peaks) decreases and goes toward a sta- ble level; it is always smaller during direct transformation (g-«-18R) than during

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1982497

(3)

C4-616 JOURNAL DE PHYSIQUE

reverse transformation (18R-B).

In the present work measurements on a BCu-Zn-Al alloy which transforms into 2H m e n s i t e have been carried out. The principal features were extracted and compared to those of BCu-Zn-Al transforming into 18R madensite because it is h o w that some growth and characteristics are different.

Experimental.-The experimental set-up was already described elsewhere (9). A small mass conduction calorimeter allows to follow relatively fast the temperature pro- grams, its response can be accelerated by an inverse analog and/or digital filter

(10). A time constant less than 0.5 s results from this correction.

A piezoelectric transducer (P.Z.T. ceramic) is bounded to one flat sample surface within the calorimeter cell. The signal given by the transducer is amplified (80 to 90 db) and filtered (bandwidth between 10 kHz and 2 MHz). It is then treated by a counting block (triggering level-1V) wh31ch allows the ring-down counting and the ring-down counting rate.

The alloy composition is Cu; 7.18 at% Zn; 23.12 at% Al. Two samples were used, the sample 1 is a 6 single crystal (15 m x 10 m x 2 m ) and was obtained by a mdif$ed Bridgmn method; the sample 2 ( 9 m x 9 m x 0.8 m) is polycristalline

(mean grain diameter 2 m) and was obtained after induction melting and hot extru- sion. They were heated during 15 min at 1120 K quenched in water at room temperature.

The martensite structure was found to betwinnedby electron microscopy and the 2H structure was identified by diffraction patterns.

Experimental results.-It must be first emphasized that no significant difference is obtained between the single and polycristalline samples. This m y be due to that large grain size of the polycristalline sample.

Figure l(a) shows the simultaneous recorded spectrm. of the thermal power and A.E. (ring-down count rate) of the entire forward martensitic transfomtion.

For comparison the same type of record, corresponding to the 6-18R transfomtion, is displayed on fi'pe l(b).

Records of two different cycles are presented on f i p e s 2. The thermal and acoustic recordings have the same behaviom: the continuous background signal is very weak whereas the peaks are very sharp and high. The transformation proceeds therefore very discontinuously ("jerky character").

(4)

Fig. l(a): General behaviour of A.E. and thermal power dissipated during direct martensitic transformation. Sample l. dT/dt = 0.0022 Ks-' . ( ) For comparison the same recordings for @--18R transformations. dT/dt = 0.0025 Ks4.

The thermal and acoustic peaks are very well correlated as shown in figure 3 which is a detailed prt of the s p e c t m recorded during the forward transformation.

From one transformation cycle to the other the effects are not reproducible, meaning that the peaks are not located at the same temperatures, but the transformation shows up the same very discontinuous behaviour (figure 4 ) .

Tempera ture (K)

Fig. 2:Details of records for the more active part of the forward transformation and corresponding to two different transfomtion cycles. Sample 1. dT/dt= 0.0022Ks-'

The cumulative A.E. changes from one transformation cycle to the other but without clear evolution (see figures 5 a and b). Due to the discontinuous character of the transformation the transition temperature (Ms point) is difficult to locate.

The sample studied seems to have a temperature range within which the peaks are ve-

(5)

C4-618 JOURNAL DE PHYSIQUE

ry high and very dense (fig. 3) (sometimes it can be evaluated that half the sample has been transformed in that domain). The mean temperature of this temperature range changes from cycle to cycle (fig. 4 ) .

Temperature (K)

F+g.,3: Recording showing the very good correlation between A.E. and thermal power dlsslpatlon (detail). Sample 2.

Temperature (KI Number of cycles

Fig. 4 :Uncorrected them-ogr- (without Fig. 5 : Variation with the number of trans- s s efilter which gives an integra- formation cycles of the relative cumulated tion of the peaks) corresponding to the A.E. (Sample 1 (a) and sample 2(b) )

.

four first B-2H transformation cycles

of s m l e 2. The initial transformation B--2H 2H--6 zone is very active and fluctuates in

temperature. The lack of reproducibility is clearly visible.

In order to better understand this lack of reproducibility in the thermal and acoustic recordings, the formation of martensite plates was followed by microscopy

(surface relief) on the same samples. It is observed that the martensite variants

(6)

The inverse transformation has a similar behaviour if compared to the forward transformation, with the exception that the continuous background seems to be relatively large with respect to the peaks. This reverse transformation seems therefore to be less discontinuous than the fornard transformation (fig. 7 ) . However it may be impoi3ant to notice that the cumulated A.E. generated during the reverse transformation is more or less the same as *he one generated during the forward transformation.

This is in opposition to the observations in 6 Cu-Zn-A1 alloys exhibiting B e 1 8 R transfomtion where the A.E. is always 2 to 5 times higher during the reverse transformation if compared to the forward one (6). Moreover the temperature hysteresis is much higher in the case of 6e2H (Af - Ms -18 K) than in the case of B=18R transformations (Af - Ms

-

^{5 K)}

^.

Fig. 6 : Optical micm,qaphs (surface relief) taken at the same sample and corresponding to two successive cycles (40 and 41 cycles).

Fig.7: General behaviour of the A.E. and thermal power recording for the reverse transformation. Sample 1.

263 261

Temperature (KI

Conclusion.- From simultaneous measurements of A.E. and thermal power, several characteristics concerning the 6 ~ 2 H martensitic transformation in a Cu-Zn-Al alloy

(7)

JOURNAL DE PHYSIQUE

have been analysed.

The transformation has very clear jeregmwth character behaviom and this be- haviow does not evolute for successive transformation cycles in contrast to what it is observed for the Bt'18R transformation. The transformation appears not to be reproducible if successive cycles are compared. This is confirmed by microstructural observations. This lack of reproducibility is probably related to the discontinuous character of the transfomtion (jerky behaviour). The cumulated A.E. which is different from one transformation cycle to the other comborates this idea.

All these features seems to question the thermoelastic growth of this transformation in spite of a relatively small temperature hysteresis (Af - Ms

-

¹⁸^{K ) .}

In particular it has been shown that nearly all the energy relased is contained in the sharp peaks, which correspond to large acoustic emission. This pmves that slow moving interfaces are probably in minority or even non-existent.

Acknowledgements.- A.P. gratefully achowledges aidfmm Centre International des Etudiants Stagiaires (France) for a scientific stay at Laboratoire de Traitement du Signal et Ultrasons, I.N.S.A., France.

References. -

1. G.R. Speich, R.M. Fischer, Acoustic Emission (AmericanS~ciet~for Testing Mate- rials) STP 505 (1972) 140

2. R. Pascual, M. Ahlers, R. Rapcioli, F?. Arneodo. Scripta Metall.

9

(1975) 79

3 . J. Baran, M. Rosen. ~hil. Mag. A

44

(1981) 895

4. E. Esmail, I. Grabec, V. KraSevec. Proceedings of Ultrasonic International 79, Graz, Austria 1979, IPL Science and Technology, p.45

5. C.H. &ceres, W. Arneodo, R. Pascual, H.R. Bertorello. Scripta Metall.

14

(1980) 293

6. A. Planes, J.L. Macqueron, M. Morh, G. Gugnin. Phys. Status Solidi (a) 66 (1981) 717

-

7. J. L. Macauemn, P. Fleischmnn, H. Moulinier. Proceedings of Journges de Calori- mgtrie et Analyse Thermique, AFCAT, Paris 1977 p. 281

8. H. Warlimont, L. Delaey. Progress in Material Science eds. B. Chalmers, J.W.

Christian, T. B. Massdlski

,

^V.18, Pergamon Press, Oxford 1974

9. A. Planes, J. L. Macquemn. Proceedings of Journges de CalorimStrie et Analyse Thermique, AFCAT, Barcelona 1980 p.1.7.1.

10. E. Cesari, V. Toma, J.L. Macquemn, R. &st, J.P. Dubes, H. Tachoire. Thermo- chim. Acta

53

(1982) 1