THE INFLUENCE OF MARTENSITE STABILISATION ON CHANGES IN DISLOCATION DENSITY IN Cu-Zn-Al ALLOY

HAL Id: jpa-00227191

https://hal.archives-ouvertes.fr/jpa-00227191

Submitted on 1 Jan 1987

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 INFLUENCE OF MARTENSITE

STABILISATION ON CHANGES IN DISLOCATION DENSITY IN Cu-Zn-Al ALLOY

J. Ilczuk, L. Delaey, J. van Humbeeck

To cite this version:

J. Ilczuk, L. Delaey, J. van Humbeeck. THE INFLUENCE OF MARTENSITE STABILISATION

ON CHANGES IN DISLOCATION DENSITY IN Cu-Zn-Al ALLOY. Journal de Physique Colloques,

1987, 48 (C8), pp.C8-553-C8-558. �10.1051/jphyscol:1987887�. �jpa-00227191�

au

THE INFLUENCE OF MARTENSITE STABILISATION ON CHANGES IN DISLOCATION DENSITY IN Cu-Zn-A1 ALLOY

J. ILCZUK, L. DELAEY* and J. VAN HUMBEECK*

Institute of Physics and Chemistry of Metals, Silesian University, Bankowa 12, PA-40-007 Katowice, Poland

" ~ e p a r t m e n t of Metallurgy and Materials Engineering, Katholieke Universiteit Leuven, de Croylaan 2, B-3030 Eeverlee-Leuven, Be1 gium

Resume

-

On a examine l'alliage Cu-Zn-A1 avec l'effet de memoire de forme, utilisant la methode du frottement interne. En Btudiant les dependances d'amplitude du frottement interne on a constate que pendant la stabilisation de la martensite la temperature ambiante, la densite et la mobilite des dislocations augmentent la presence des lacunes trempBes.

Abstract

-

From analysis of the amplitude dependence of internal friction for Cu-Zn-A1 alloy exhibiting the shape memory effect, the changes taking place in the structure of the martensite during ageing at room temperature were determined.. It was ascertained that during stabilisation of the martensite, both the density and mobility of dislocations increase due to the presence in the martensite of a substantial concentration of quenched-in vacancies.

I

-

INTRODUCTION

The practical application of Cu-Zn-A1 alloys exhibiting the shape memory effect associated with reversible martensitic transformation is governed by the condition of ensuring complete shape recovery and stability of transformation temperatures.

Considerable attention has recently been devoted to the stability of properties and structure, and also to the understanding of the mechanism responsible for the process of stabilisation of these alloys, employing such research methods as : electron microscopy, X-ray diffraction. microcalorimetric measurements, measurements of mechanical properties and also of internal friction (IF) [l

-

^101.

From investigations already carried out it was ascertained that the process of martensite stabilisation is a complex thermally activated process, markedly dependent on the concentration of quenched-in vacancies formed during quenching from a high temperature, leading to an increase in the temperatures of transformation, relative to those for nonstabilised alloys. The object of the investigations reported here was to study the influence of the process of martensite stabilisation in a Cu-(14.5%wt.)Zn-(8.5%wt.)Al alloy on variations in density and mobility of dislocations arising during this process taking place at room temperature, by making use of measurements of IF-amplitude dependence.

I1

-

TEST MATERIALS AND METHODOLOGY

Tests were performed on samples of Cu-Zn-A1 obtained by melting copper, zinc and aluminium of technical purity in an induction furnace. The ingot was homogenised by annealing at 1220K for 12 hours and then hot-rolled (970K) to a thickness of 0.8 mm. Samples of dimensions 35~10~0.8 mm were cut from the sheet obtained and used for IF measurements. To achieve a martensitic structure, samples were heated

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1987887

C8-554 JOURNAL DE PHYSIQUE

a t 1020K f o r 15 minutes, next quenched t o 573K a t which temperature they were held f o r 15 seconds and then quenched i n ice-water. Measurements of I F amplitude dependence were conducted a t room temperature using a DMA (Dynamic Mechanical Analyser) acoustic frequencies relaxator, delivered by the DUPONT company [6] i n a range of vibration amplitudes from 3-10-5 t o 6-10-', both immediately following heat treatment and a l s o a f t e r various ageing times a t room temperature. Parallel with measurements of curves of Q-l = f ( & ) , resonance frequency of sample vibrations was a l s o measured. Based on these measured values i t was possible t o determine variations i n Young's modulus, which i s proportional t o the square of the frequency.

I11

-

MEASUREMENT RESULTS AND DISCUSSION

Fig. 1 shows the curves of i n t e r n a l f r i c t i o n and f 2 a s a function of measurement time. measured a t room temperature f o r a sample immediately a f t e r heat treatment.

t [min.]

I 1 _b

20 10 GO

Fig. 1 Internal f r i c t i o n and square of resonance frequency of the sample as a function of time of ageing a t room temperature (T = 293K) ; E = 3.2-10-5

On the Q-I = f ( t ) curve may be seen a d i s t i n c t i n t e r n a l f r i c t i o n peak ( t = 2 min) and a corresponding reduction i n value of Young's modulus a f t e r which the modulus value increases with increase i n ageing time. This observed I F peak i s known as the

"peaking e f f e c t of the f i r s t kind" (PEI) and occurs due t o the interaction between vacancies and dislocations a t the boundaries of the martensite p l a t e s [7,10]. With the aim of studying the changes taking place during s t a b i l i s a t i o n of the martensite, curves of IF-amplitude dependence were determined a f t e r various ageing times. A s an example, on Fig. 2 a r e shown the curves Q-l = f ( & ) and f 2 ( 8 ) obtained d i r e c t l y a f t e r heat treatment (curve a ) and also a f t e r 10 min ( b ) , 30 min (c) and 60 min ageing ( d )

.

stabilisation times at room temperature :

a,a'

-

measurement immediately after heat treatment.

b,bl

-

after 10 min stabilisation.

c,c'

-

^{30 minutes}

d,d'

-

60 minutes

On the curves Q - I = f(8) may be observed initially a drop in IF, after which in the range of vibration amplitudes from 5-10-5 to l - l ~ - ~ a small IF maximum is found whose height rapidly decreased during 20 minutes and next rose slightly (Fig. 3).

Fig.

3

Height of IF amplitude peak as a function of sample ageing time at room temperature

C8-556 JOURNAL DE PHYSIQUE

For &>1.10-~ a zone of IF weakly dependent on amplitude was observed, which increased with increasing ageing time, after which IF showed a rise. An IF maximum occurring on the IF amplitude curves has also been observed for, among others, pure aluminium samples containing a slight quantity of copper or magnesium, subjected to cold plastic deformation [ll]. In the initial phase of the martensite stabilisation process the vacancies migrate to the dislocations, causing them to be pinned. The concentration of dislocation pinning points is not very large so that they can easily be unpinned which leads to a slight rise in the height of the IF peak as the ageing time is prolonged. For short ageing times the value of Young's modulus shows a slight increase, which is a corroboration of the postulated mechanism of formation of this peak in the IF-amplitude dependences. With increase in length of ageing time a weak increase in modulus value may be observed.

The IF-amplitude dependence fusing analysed was the Granato-LUcke model. Fig. 4 shows the curves of In( Q-'-&) as a function of ' / E obtained from the Q-' = f(&) curves given on Fig. 2 , for a range of vibration amplitudes from 2 - l 0 - ~ .to 6.10-~.

It may be seen that the points lie along straight lines, which is a confirmation that the high damping in the martensite phase is related to dislocations.

Fig. 4 Granato-Liicke plots obtained from the IF-amplitude dependence curves shown on Fig. 2

Then knowing the parameters of the Granato-LUcke curves and the strain amplitude 8 ,

corresponding to the stress value needed to tear the dislocation segments away from the strong pinning points (L,), using the relation &, = b/LN [I21 (where b is the Burgers vector). the change in dislocation density as a function of time was estimated. On Fig.

5

(curve a) are shown the changes in dislocation density Ai/Ao

(A, is the dislocation density observed after 5 minutes ageing at room temperature, counting from the moment of finishing heat treatment) taking place during the process of martensite stabilisation. It may be seen that during 40 minutes ageing time the dislocation density does not change, after which it rises, achieving after 60 minutes a level 1.5 times that observed after

5

minutes. Increase in dislocation density in the martensitic phase is also observed after cyclic martensitic transformations [13

-

^161.

Using the measured values of vibration frequency f and knowing the variations of Young's modulus AE/ E, and also dislocation density and value 8 , . the dislocation mobility was estimated, using the following relation [IT] :

A A A

13 -

12

-

11

- 14

13

12

-

-

-

a b c

-

13

-

12 -

11

stabilisation at room temperature are shown on Fig. 5 (curve b).

Fig.

5

Variations in density (curve a) and mobility (curve b) of dislocations, taking place during martensite stabilisation at room temperature

After 40 minutes ageing mobility rises to 4 times its initial value. and next begins to decrease reaching after 60 minutes a value 3.5 times greater than that found after

5

minutes. The observed changes in both density and mobility of dislocations suggest that the vacancies frozen due to step-quenching from a high temperature, during the ageing process form clusters leading to the formation of dislocation loops and stacking faults. Stabilisation of the martensite is governed both by pinning of the martensite boundaries and dislocations by the migrating vacancies and also by the formation of stacking faults.

IV

-

CONCLUSIONS

-

Analysis of IF-amplitude dependences showed that the martensite stabilisation process in Cu-Zn-A1 alloy exhibiting the shape memory effect, is associated with freezing of the saturated vacancies in the martensite.

-

During the martensite stabilisation process the quenched-in vacancies form clusters which are, presumably, the cause of the formation of dislocation loops and stacking faults. which are responsible for the variations in density and mobility of dislocations.

-

^After 60 minutes ageing at room temperature a 1.5 times growth in density and a 3.5 times growth in mobility of dislocations is observed, relative to the values found immediately after heat treatment.

REFERENCES

[I] Van Humbeeck J., Jansen J., Mwamba N. and Delaey L., Scripta Met., l8, (1984), 893

[2] ne1aey L., Suzuki T. and Van Humbeeck J., Scripta Met.,

18,

^{(1984). 899}

[3] De Graaf M., Van Humbeeck J.. Andrade M. and Delaey L., Scripta Met.,

9,

(19851, 643

[4] Scarsbrook G., Cook J.M. and Stobbs W.M., Met. Trans.,

m,

(1984). 1977 [5] Abu-Arab A.. Chandrasekaran M. and Ahlers M., Scripta Met.,

18,

(1984), 709 [6] Van Humbeeck J

.

and Delaey L. , Z. Metallkunde,

E ,

(1984) , 755

[7] Van Humbeeck J. and Delaey L., Z. Metallkunde.

E ,

(1984), 760 181 Kondratiev S.Ju., Jaroslavskii G.Ja.. Koval Ju.N., Chaykovskii B.S.,

Zotov O.G. and Choroshaylov B.G.. Fiz. Met. Metall., B, (1985). 160

C8-558 JOURNAL DE PHYSIQUE

[g] ^{Bojarski Z., Ilczuk} J., Panek T. and Morawiec H., J. Physique,

44,

(1983), C9-241

[lo] Van Humbeeck J . , Hulsbosch A,, Delaey L. and De Batist R., J. Physique,

46,

(1985). ClO-633

[ll] Ke T.S., J. physique,

46,

(1985). C10-267

[I21 Krishtal M.A., Golovin S.A. and Arkhangelskii S.I., Fiz. Met. Metall., 21, (1966). 83

[I31 Perkins J.. Met. Trans.,

4,

(1973). 2709

[14] Kajiwara S. and Kikuchi T.. Acta Met..

9 ,

(1982), 589

[15] Bojarski Z., Morawiec H. and Panek T., Phys. Stat. Sol. (a),

81,

(1984), K97 [I61 ^Ilczuk J. and Morawiec H., Phys. Stat. Sol. (a),

z,

(1986). 371

[ I T ] Baker G.S.. J. Appl. Phys., 3 , (1962), 1730