REVIEW: COPPER BASED SHAPE MEMORY ALLOY …ijirae.com/volumes/vol1/issue8/SPME10080.06.pdf · Madikeri, Karnataka, India Abstract: ... the elemental mapping images of the milled powders

International Journal of Innovative Research in Advanced Engineering (IJIRAE) ISSN:2349-2163 Volume 1 Issue 8 (September 2014) www.ijirae.com

_________________________________________________________________________________________________ © 2014, IJIRAE- All Rights Reserved Page - 38

REVIEW: COPPER BASED SHAPE MEMORY ALLOY FOR REINFORCING INTO ADAPTIVE COMPOSITES

Kotresh M

Research Student, Department of Mechanical Engineering, VTU Research Resource Centre, Belgaum,

Karnataka, India

Dr. M M Benal

Professor and Head ,Department of Mechanical Engineering, Govt. Engineering College, Kushalnagar,

Madikeri, Karnataka, India

Abstract: Incorporation of shape memory alloy materials with the composite materials can be manufactured and evaluated which may utilize the properties of individual bulk materials to achieve optimal properties. A variety of Shape memory alloy composites have been studied with shape memory elements being reinforcement. The adaptive composites provide enormous potential for creating new paradigms for material structure interaction and exploring varying success in many engineering applications. This review, from the standpoint of materials science, will provide a survey on Cu based shape memory powder developed during last decades. Emphasis placed on fabricate or manufacturing, characterization and behavioural studies of shape memory particles for composites.

Key words: Shape Memory Alloy, Superelasticity, Superplasticity, Mechanical Alloying (MA),Adaptive Composites

I. INTRODUCTION Shape memory alloys (SMA) are now conventional functional materials with intrinsic characteristics that

differentiate from other materials. Because of low price and high recovery force Cu based shape memory alloys are most promising in practical use. Among the Cu based shape memory alloys Cu-Al-Ni (CAN) has higher thermal stability than others. For the preparation of the SMA conventional casting methods have difficulties in controlling the grain size. Moreover the composition change during casting can shift the transformation temperature. The mechanical alloying (MA) technique has been used to produce dispersion strengthened alloys [1-3] and amorphous alloy powder in the solid state [4-6]. One of the advantages of MA is the ease of preparing alloy powders containing both high and low melting elements. In this process, powder are cold welded and then fractured repeatedly during the high energy collision between balls. As well as between the balls and the walls of the jar. The repeated welding and fracture finally lead to formation of alloy. It has been reported that MA and powder metallurgy (PM) can be used to fabricate Cu based SMA, which allow us to improve the mechanical and thermo-mechanical behaviour of the alloy. Therefore following aspects such as characterization, superelastic effect, superplastic effect, microstructure and thermo mechanical relationship of SMA are discussed [5].

II. EXPERIMENTAL WORKS

Structure evolution of CAN SMA during MA process Material Preparation & Performance:

In the MA process, a QM-1F high-energy planetary ball mill with four stainless steel vials was used, and each vial contains hardened steel balls of different sizes, that is, 20, 10 and 6 mm in diameter. The ball-to-powder mass ratio (BPR) is 15:1.The specification of elemental powders and the initial powder mixture is listed in Table 1. The sealed vials were evacuated and then filled with argon gas. The detailed milling processing is listed in Table 2.

A small amount of milled powder was removed after certain milling time from the container in an argon glove box and investigated using X-ray diffractometry (XRD) with Cu Kα radiation in Dmax-2500 diffractometer [1]. The morphology of the milled powders was observed using Laica EC1 metallographic microscope; the elemental mapping images of the milled powders was observed using scanning electron microscope(SEM) Sirion 200 equipped with

Table 1. Specification and mixture of elemental powders

Powder Size/µm Purity % Composition of mixture (mass fraction)% Cu 75 99.0 81.5 Al 75 99.9 12.5 Ni 75 99.9 6

Table 2. Details of milling process of mixing powder

Project Rotation speed(rev/min) Milling time/h Cu-Al-Ni 300 1,5,15,25,50



EDAX GENESIS 60 [3].The Vickers hardness of powders was measured using a load of 1 N and holding for 15 s on a Vickers hardness machine. The evolution in morphology and particle size of the MA powders with milling time is observed and shown in Fig.1. For 300r/min, the powder particles agglomerate with prolonging milling time. However, the degree of agglomerating of powder particles at 200 rev/min for 50 h is obviously lower than that at 300 rev/min for 50 h. MA involves two opposite processes: welding and fracture among powders. In general, at the early stage of milling, welding is in dominating situation [13].at this stage, the powders are easily welded, resulting in an increase in the size of powders. The fracture of the powder takes place subsequently. However, the tendency of welding and fracture depends on the properties of the powders alloyed. When soft materials are used, welding is excessive, while for the brittle materials, fracture dominates [6]. The Cu, Al and Ni powders are soft and ductile at room temperature and on high speed milling, the temperature of the powder increases; the Cu, Al and Ni powders become even softer and more ductile so that excessive welding occurs. The size of the powders is observed to increase to 0.5 mm while distribution of size is not uniform.

Figure 1. Morphology of MA Powder a.300 rev/min,5h b.300 rev/min,25h c.300 rev/min,50h d.200 rev/min,50h Hardness Measurement:

The hardness of the milled powders as a function of the MA time is given in Fig.2. Initially the hardness of the powders increases rapidly because of work hardening. The maximum hardness is reached after milling for 5 h both at 200 rev/min and at 300 rev/min; then the hardness decreases until milling for 15 h. It is believed that the decrease in hardness is attributed to the increase in temperature during MA, which causes the hardened powders to be kinetically annealed. After MA for 15 hardness increases slightly, this may be attributed to the solution of Ni and Al into Cu. However, hardness measurement in the present study is not a unique measurement to evaluate the degree of MA. For softer materials, the increase in hardness can prevent excessive welding and hence the efficiency of MA can be improved [6]. X-ray diffraction analysis: X-ray diffraction profile of powder milled for different time at various milling speeds is shown in Fig.3. That milled at 100 rev/min is shown in Figs.3(a) and (d). The pattern of 5 h MAed powders is taken as the powders of Cu, Al and Ni .

Figure 2. Vickers hardness change over milling time



Figure 3. X-Ray diffraction pattern There are no significant changes in the XRD pattern after milling for 50 h. This indicates that the energy is not enough to make these elemental powders form solid solution effectively at 100 rev/min using high energy planetary ball mill [8]. Figs.3(c)−(f) show the X-ray diffraction profile of powder milled at 200 rev/min. With the increase of milling time, the intensities of the diffraction peaks of Al decrease, and those of the Ni almost remain unchanged, which indicates that no significant reaction occurs during milling and the solution of Ni in Cu matrix is more difficult than that of Al. The diffraction peaks of Al and Ni can be seen even after 50 h of milling. It is difficult to enhance the solid solubility of the pre-alloying powders by increasing milling time if the rotation rate is 200 r/min. X-ray diffraction profiles of powder milled at 300 rev/minare shown in Figs.3 (g)−(k).The intensities of Cu diffraction peaks increase and those of Al and Ni diffraction peaks decrease with prolonging milling time [10]. Ni and Al peaks almost disappear just after MA for 15 h. The position of Cu diffraction peaks move toward slow-angle, which indicates that the lattice parameters of Cu matrix increase with milling time. The X-ray diffraction results, which agree well with the observation by KANEYOSHI et al [14], lead to the conclusion that a single phase solid solution is formed after 25 h MA of the elemental power mixtures at 300 rev/min. After further milling, the XRD pattern where the spectrum consisting of Cu diffraction peaks super imposes on top of broad background presents. This broad background is associated with amorphous phase [15].The width of peaks increases at all tested milling speeds with balling time. It is caused by the decrease of crystallite size and the increase of micro-strain due to a high stress evolved in milling balls impacts. Considering that the increase of peak width mainly consists of the Cauchy part that caused by decrease of crystallite size and Gauss part caused by increase of micro-strain, the XRD profiles can be fitted by Pearson-VII function. Then the average crystallite size D and micro-strain ε can be calculated by substituting the fitted integral breadth into D=Kᅑ/ᅈ F

c Cosϴ and ɛ=(1/4) ᅈ F G Cotϴ Where ᅈ F

c and ᅈ F

G are the Cauchy’s and Gauss integral breadth. The constant K is 0.9 and λ is 0.15405 nm. The changes in crystallite size and micro-strain of Cu matrix as a function of milling time are shown in Fig. 4. The crystallite size decreases while the micro strain increases with balling time. The faster the rotation Speed, the larger the changes of crystallite size and micro-strain. The Cu powders have an average crystallize size of 10 nm after milling for 50 h at 300 rev/min.

Figure 4. Change of Crystalline size



Microstructure analysis: Fig.5 shows the microstructure evolution of the MA powders with milling time at 300 rev/min. The laminated structure in the powder particle is seen by plastic deformation and cold welding, and the layer becomes thin with the increase of milling time. The layer is thick but their thickness is not homogeneous after milling for 5 h (Fig.5 (a)). With the increase of milling time, the laminated structure is discontinuous, and the thickness of the layer decrease, as shown in Figs. 5(b) and (c).The Cu-matrix becomes homogeneous as the milling time approaches 50 h, as shown in Fig. 5(d). However, the layered structure of powders can be clearly observed at 200 rev/min for 50 h (Fig. 5(a)). The SEM images and the elemental mapping images of the MA alloyed powders at 200 and 300 rev/min for 50 h are shown in Fig.6. The layered structure of Cu, Al, Ni powders can be obviously observed at 200 rev/min for 50 h (Fig. 6(a)). The elemental mapping images taken from Fig. 6(a) are shown in Figs.7(c), (e) and (g). Cu, Al, Ni elements distribute non-uniformly in a powder particle, which illuminates that the alloying of powder mixture is not enough [13]. At 300 rev/min for 50 h, the microstructure is shown in Fig. 6(b), and the layer structure is thin and only exists in partial area. The elemental mapping images taken from Fig. 6(b) are shown in Figs. 7(d), (f) and (h). Cu, Al, Ni elements distribute uniformly in a powder particle, and this suggests that the alloying of powder mixture is enough processed [7-10]. This observation agrees well with the XRD analysis mention above. The Cu-matrix becomes homogeneous.

Figure 5. Typical laminated structure after MA

Figure 6. SEM Images of MA powder from 200 to 300 rev/min.



Superplasticity of Cu-Al-Mn-Ni SMA Material Preparation: Cu71.5Al17Mn11.5 and Cu71.3Al17Mn8.3Ni3 alloys were prepared by induction melting under an argon atmosphere. The ingots were hot-rolled to a thickness of 10 mm at 800oC and subsequently cold-rolled to a thickness of 1.15 mm with intermediate annealing at 600oC. Specimens were cut from the sheets, and heat treatment was conducted at 600oC for 30 min or at 900oC for 15 min. The microstructure was observed using an optical microscope and the volume fraction of the phase was determined by image analysis [5-12]. Tensile tests were carried out by an Instron-type machine at 450oC and 500oC, where the tensile axis was parallel to the rolling direction of the sheet and the strain rate was constant in the range of 1x10-1s-1-5x10-4 s-1.

Material Performance: Microstructure of Cu-Al-Mn and Cu-Al-Mn-Ni alloys From Fig. 8, it can be expected that the ductile Cu-Al-Mn shape memory alloy with 17 at% Al shows the ᅈ single phase at temperatures over about 700 oC, while the α+β two-phase structure appears at lower temperatures. Since the phase is very ductile due to the disordered fcc structure, improvement of the ductility for the Cu-Al-Mn-base shape memory alloys is possible by the introduction of the α phase in the ᅈ phase. Therefore the annealing temperature suitable for subsequent cold-working is around 600 oC and the final shape memory treatment to obtain the β single phase can be performed at temperatures over 800 oC. Figures 7(a) and (b) show the single-phase microstructure with a grain diameter of 400–500 µm annealed at 900 oC for 15 min and the α+β two-phase structure annealed at 600 oC for 30 min in the Cu71.5 Al17

Mn11.5 alloy, respectively [15]. The optical micrograph in Fig. 7(c) shows the α+β two-phase microstructure taken from the Cu71.3Al17Mn8.7Ni3alloy annealed at 600 oC for 30 min, which exhibits a much finer two phase microstructure with a mean grain diameter of approximately 3 µm. The volume fractions of the α phase for the ternary and quaternary alloys shown in Figs. 7(b) and (c) are 42% and 64%, respectively. This result is consistent with the phase diagram shown in Fig. 9 between the and phases at 700oC, where the partition of Mn between the α and ᅈ phases is almost equal. It is seen that the α+β two-phase region is widened by the addition of Ni and that the volume fraction of the α phase increases with the addition of Ni at Al content of 17 at% [11]. The increase of the phase volume improves the cold workability, and the maximum reduction in thickness before a crack appears in a sheet by cold-rolling increases from 60% to 80% by the addition of Ni. The Cu-Al-Mn-Ni shape memory alloy with the α+β two phase microstructure, which was obtained by annealing at 600 oC, seems to have a microstructure suitable for obtaining Superplasticity, i.e., the grain size is less than 10–15 µm.

Figure 7. Microstructure of Single Phase and two phase Cu-Al-Ni. Figure 8. Vertical section of 10 at% Mn in Cu-Al-Mn system with phase of Cu-Al binary system indicated by broken line.



Superplasticity: The tensile test was carried out for the Cu71.3Al17Mn8.7Ni3 alloy annealed at 600 oC for 30 min, whose microstructure is shown in Fig. 7(c). Figures 9(a) and (b) show the stress strain curves at 450oC and 500 oC, respectively. Figure 10. shows the elongation to failure as a function of strain rate. The elongation is enhanced several hundred percent by deformation at a lower strain rate and a higher temperature; in particular, that at 5x10-4s-1and 500 oC reaches 1150% [4]. Figure 11. shows the flow stress against strain rate. It is seen that the flow stress increases with increasing strain rate and that at the same strain rate, the flow stress at 450 oC is always higher than that at 500oC. The strain rate sensitivity, m, at 450 oC is about 0.37 and that at 500 oC is over 0.33. This plastic behaviour with the large tensile elongation prior to failure and the large strain rate dependence of flow stress is clearly due to Superplasticity. Actually, it has been reported that the Cu-Zn superplastic alloy with a α+β two phase microstructure shows m=0.1-0.2 at 400 oC and m=0.2-0.4 at 500 oC at strain rates of 1x10-3s-1–1x10-2s-1.It can be concluded that the Cu-Al-Mn-Ni two phase alloy, which is similar to other superplastic Cu-base alloys, is a superplastic shape memory alloy [16].Figure 10. Displays the specimens before and after the tensile test. While necking can be observed in the specimens deformed at higher strain rates at the lower temperature, the specimens at lower strain rates at the higher temperature show a large elongation without necking [12]. Figure 9. Stress-strain curves of Cu713Al17Mn87Ni3alloy at indicated strain rates at (a) 450 oC and (b) 500 oC.

Figure 10. Shows the elongation to failure as a function of strain rate.

Microstructure after tensile test Figures 12(a) and (c) show the microstructure near the grip of the tensile test specimens, and (b) and (d) are taken from the regions near the fracture tip [8-13]. Here, the cavities are seen lying parallel to the tensile axis and the shape of grains is approximately equiaxed as shown in Figs. 12(b) and (d), which are common features in the superplastic alloys. It is also observed in the specimen after the 1150% elongation (in Fig. 12(d)) that the mean size of the cavities is larger than that of the specimen subjected to static isothermal annealing (in Fig. 12(c)) and that noticeable grain coarsening occurs.



This is clear evidence of strain-enhanced grain growth due to the high degree of deformation, which is consistent with conventional knowledge that the magnitude of grain growth increases with increasing strain [17-21].

Figure 11. Shows the flow stress against strain rate Relationship between micro structure and thermo mechanical properties of Cu-Al-Ni shape memory alloys obtained by powder metallurgy: An alloy of composition Cu-14.2Al-4.2Ni (% wt.) has been obtained by the PM processing method presented above. The microstructural characterization of FP1 and FP2 has been carried out by means of SEM in a JEOL 6400 with EDX and WDX microanalysis, and TEM in a Philips CM200 with EDX microanalysis, operated at 200 kV using a double tilt holder [4-12]. Material Performance: Figures 13a and 13b show the microstructure of FP1 and FP2 observed by means of SEM in BSE mode. In FP1 the powder particles used for compacting can still be seen. They are spherical with a medium size of about 50 µm, rounded

Figure 12. Microstructures of specimens grip tensile-tested

by a thin oxide film coming from the surface of the particles [13-20]. This film, which has been carefully studied by EDX microanalysis, stops grain growth during compacting and prevents cohesion between the particles, being the responsible of the sample brittleness. After hot rolling, the oxide film gets broken up, giving a final product with a higher cohesion between the particles and a higher texture which has been determined in the SEM by Electron Channelling Pattern (ECP) and Electron Backscattered Diffraction (EBSD) [14]. FP2 also presents a dislocation structure in which poligonyzed subjoins can be observed, whereas they were not observed in FP1. As it was impossible to study these subjoins by bright field (BF) imaging due to the difficulty to decide if they were composed by dislocations or by super dislocations, weak beam (WB) was used in order to clarify the core structure of the dislocations. As it can clearly be seen in Figure 13c, these subjoins are formed by super dislocations, leaving an antiphase boundary between them, which seems to be logical in an ordered structure [16-19].After martensitic transformation, martensite domains appear and get propagated across the subjoin structure (Figure 13d) without being affected by its presence [20]. In this way, hot rolling



has generated a new microstructure that improves the ductility without affecting the martensitic transformation behaviour of the produced alloy.

Figure 13. a) SEM Micrographs showing microstructure of FP1 b) SEM Micrographs showing microstructure of FP2 c) Super location subjoins in FP2; d) Martensite domains crossing subjoins in FP2 (BF, g=220)

III. CONCLUSION

Numerous smart hybrid composites can be designed by incorporating shape-memory material particles with other advanced materials. The shape memory hybrid composites show some unique properties or which can be utilized to tailor or tune the overall performance of a smart structural system. Thanks to the various efforts and exploratory slants being made by many researchers, during the past span the shape-memory hybrid composites have progressed and become one of the promising advanced composites for smart systems. However, the research and growth of the smart composites is still in its early stage; many problems remain unresolved and some technical challenges lie ahead. REFERENCES: 1.Toshihiro Omori and Naoki Koeda in “Journal of Japan Research Institute for Advanced Copper-Base Materials and

Technologies” Materials Transactions, Vol. 48, No. 11 (2007) pp. 2914 to 2918 2. P. P. Rodríguez, R. B. Pérez-Sáez in” EUREM 1”2, Brno, Czech Republic, July 9-14, 2000 3. XIAO Zhu and LI Zhou in Trans. Nonferrous Met. Soc. China 17(2007) 1422-1427 4.C.A.ROGERS, in ‘‘Smart Materials, Structures, and Mathematical Issues’’, edited by C.A. Rogers (Technomic,

Lancaster, 1989) p. 221. 5. C. A. JAEGAR and C. A. ROGERS, ibid., p.14. 6.B.Z.JANG, in ‘‘Proceedings of the International Conference on Advanced Composite Materials’’, edited by T.

Chandra and A. K. Dhingra (TMS, Warrendale, 1993) p. 661. 7. M.V.GANDHI and B.S.THOMPSON, in ‘‘Smart Materials and Structures’’ (Chapman and Hall, London, 1992) p.70. 8.C.A.ROGERS, in ‘‘US—Japan Workshop on smart/Intelligent Materials and Systems’’, edited by I. Ahmad et al.

(Technomic, Lancaster, 1990) p.11. 9.Idem, in ‘‘Proceedings of the American Society of Composites 3rd Technical Conference on Composite Materials’’

(Technomic, Lancaster, 1988) p.719. 10.C.C.Hsu and W.H.Wang: Mater. Sci. Eng. A 205 (1996) 247–253. 11.H.Honda, R.Matubara, N. Ashie, K. Nakamura and S. Miura: Mater. Sci. Forum 327 (2000) 477–480. 12.S.Miura:Mater.Sci.Forum394(2001)399–402. 13. X. J. Liu: Doctoral Thesis, Tohoku University (1998). 14.O.A.Kaibyshev: Superplasticity in Alloys, Intermetallides and Ceramics (Springer-Verlag, Berlin; Tokyo, 1992) pp.4–

29. 15.O.D.Sherby and J.Wadsworth: Prog. Mater. Sci. 33 (1989) 169–221. 16.J.San Juan, R.B. Pérez-Sáez, V. Recarte, M. L. Nó, G.Caruana, M.Lieblich, O.Ruano. J.de Phys. IV 5 (1995) 919. 17. J.San Juan, R. B. Pérez-Sáez, V. Recarte, O. A.Ruano, M.L.Nó. Adv.Eng. Mat. (1999) In press. 18.T.Tadaki: Shape Memory Materials, ed. by K. Otsuka andC.M.Wayman,(Cambridge University Press, Cambridge,

UK, 1998)pp. 97– 116. 19.R.Kainuma, N. Satoh, X. J. Liu, I. Ohnuma and K. Ishida: J. Alloys Comp. 266 (1998) 191–200. 20.R.Kainuma, S. Takahashi and K. Ishida: J. JRICu 34 (1995) 213–219. 21.R. Kainuma, S. Takahashi and K. Ishida: J. de Phys. IV 5 C8 (1995) 961–966.

Documents

REVIEW: COPPER BASED SHAPE MEMORY ALLOY …ijirae.com/volumes/vol1/issue8/SPME10080.06.pdf · Madikeri, Karnataka, India Abstract: ... the elemental mapping images of the milled powders