In this work, the CuAlBe shape memory alloy (SMA) with a new composition of 77.05Cu-22.02Al-0.93Be (at.%) detected by the EDS test was fabricated in a vacuum arc melter under a pure argon atmosphere. At the beginning, the high-purity (%99.9) elements of Cu, Al, and Be powders were mixed and the mixture (~ 10 g) was formed as pellets by pressure. By melting the pellets in arc melter, the as-cast ingot alloy was obtained. Then, the ingot alloy was cut into small pieces (~ 30–50 mg) and then all of these samples were all homogenized in the β-phase region (at 900 °C for 1 h) and immediately submerged in traditional iced-brine water to form β1’ martensite phase in the alloy which results from such fast cooling by suppressing the formation of hypoeutectoid precipitations (α and γ2). Then, to uncover and evaluate the existence of martensite structure and the characteristics of shape memory alloy properties of the alloy, the specimens were tested by calorimetric and structural measurements. The results of thermal heating/cooling cycles of the alloy were obtained from differential scanning calorimetry (DSC) measurements that were taken twice at 5 °C/min of heating/cooling rate under constant argon gas flow (100 ml/min) and by using liquid nitrogen cooling support to reach lower temperatures than room temperature. The DSC thermograms of the alloy revealed the characteristic martensitic transformation peaks that occurred endothermic by heating and exothermic by cooling at moderate temperatures ranging between 19 and 66 °C, regarded as evidence for the presence of shape memory effect property in the alloy. Important thermodynamical parameters such as transformation temperatures, hysteresis gap, and entropy and enthalpy change amounts for these back and forward martensitic phase transition peaks were determined directly by using data of DSC peak analyses and by calculation. Differential thermal analysis (DTA) measurement that was taken from room temperature to 900 °C at a heating/cooling rate of 25 °C/min displayed a high-temperature behavior of the alloy as compatible with the common behavior of Cu-Al SMAs. The X-ray test of the alloy conducted at room conditions showed sharp diffraction peaks and their matching crystal planes which indicate the existence of β1′ martensite phase in the alloy and its high single-like crystallinity i.e. its large Debye-Scherrer sub-micrometer crystallite size. Besides, theoretical forecasting about the existence and volumetric dominance of martensite phases was deduced from the calculated value of e/a (average conduction electron concentration per atom) parameter of the alloy. Furthermore, the mechanical Vickers microhardness tests that were performed at room temperature and under 100-gf load applied for 10 s revealed the high ductility and softness features of the alloy. Considering all the results, the highly ductile CuAlBe alloy with new composition owing shape memory alloy properties and intermediate working temperatures can be useful in various kinds of research and applications in which Cu-based SMAs are exploited.
Shape memory alloys (SMAs) are a very useful smart material class, thanks to their unique shape memory effect (SME) and superelasticity (SE). The shape memory effect (SME) property, which is made possible by reversible first-order and non-diffusional martensitic phase transformation and triggered by external factors such as temperature and pressure, can be a unidirectional (one way) or bidirectional (two way) memory feature of deformed material. It is based on the principle that it can partially or completely return to its previous shape as a result of this transformation . Shape memory alloys have several important parameters that are characterized by their SME property. The most important ones are the hysteresis interval and martensitic transformation temperatures. The shape memory effect and superelasticity of shape memory alloys take place at certain temperature ranges, thanks to martensitic transformations. Therefore, these temperature ranges are very important for the application in which the material is used . The commercial NiTi SMAs are the most preferred in the applications due to their superior SMA properties. However, these alloys have high costs and hard fabrication process disadvantages. Therefore, Cu-based SMAs (such as CuAlMn, CuAlNi, and CuAlBe), seen as the nearest alternative to NiTi ones, are widely studied by researchers due to their low cost, high strength, and high electrical and thermal conductivity properties. The advantages of copper-based SMAs made them preferred in many industrial applications from healthcare to the space industry [3, 4].
Binary Cu-Al (e.g., CuAlMn, CuAlNi, CuAlBe) SMAs are Cu-based SMAs are the most focused on . Among these, CuAlBe alloys are of interest in the intermediate and low-temperature SMA applications. By adding Be content in CuAl-based alloy, the martensitic transformation temperatures and eutectoid point decrease  without a change in the concentration [6, 7] of the eutectoid point, and this sustains the thermal stability. This also enables alloys to be designed with a wide range of transformation temperatures. Besides, lowered transformation temperatures enable the superelasticity behavior to occur at low temperatures, too.
In this study, the CuAlBe alloy with a novel unprecedented composition of 77.05Cu-22.02Al-0.93Be (at.%) was fabricated by arc melting technique. To see the effect of the minor addition of Be element on the base Cu-Al alloy matrix and to reveal the characteristic thermal and structural SME and mechanical hardness/ductility properties of the alloy, several kinds of measurements including DSC, DTA, XRD, EDS, optical microscopy, and Vickers microhardness tests were carried out.
In this study, the shape memory alloy with new 77.05Cu-22.02Al-0.93Be (at.%) was fabricated by an Edmund Bühler AM model vacuum arc melter under pure inert argon gas plasma atmosphere. Before casting the alloy in an arc melter, the high-purity (%99.9) elements of Cu, Al, and Be powders were mixed, and then this powder mixture (~ 10 g) was pressed to make pellet forms by using a hand press machine. Afterward, these pellets were together melted in an arc melter under argon atmosphere and the as-cast ingot alloy was obtained. The ingot was cut into small tablet-like pieces (~ 30–50 mg) and to homogenize the alloy (i.e. to distribute the atoms of alloying elements more ordered in the alloy matrix) all of these specimens were all solution-treated at 900 °C for 1 h in an oven. Right at the end of this homogenizing process, they were all quenched in traditional ice-brine water solution to surpass the formation hypoeutectoid precipitations (α and γ2) and by this means, the martensite phase was spawned in the alloy matrix; i.e., the shape memory effect was installed in these CuAlBe alloy samples. To observe the martensitic phase transformations in the alloy, DSC (differential scanning calorimetry) measurement by using a Shimadzu DSC-60A instrument was taken under constant argon gas flow (100 ml/min) repeated two times: first normal and then with liquid nitrogen cooling support (to reach the temperatures under room temperature) and both cycles were run at the same heating/cooling rate of 5 °C/min. To reveal the alloy’s high-temperature behavior, the DTA (differential thermal analysis) test was made by using a Shimadzu DTG-60 AH instrument at a heating/cooling rate of 25 °C/min from room temperature to 900 °C under the same constant argon gas flow. The chemical composition (at.%) of the alloy was detected by using a Bruker Model EDS (energy dispersive X-ray) instrument at room temperature. XRD test was made by using a Rigaku RadB-DMAX II diffractometer with CuKα radiation at room temperature. The optical micrographs of the alloy surface by using an XJP-6A model optical microscope with a × 250 magnification ratio were also taken at room temperature. Moreover, the Vickers microhardness test on CuAlBe alloy was also performed with indentation distancing of 200 μm horizontally and 100 μm vertically at room temperature and under 100-gf load applied for 10 s by using a Tronic DHV 1000 model Digital Microhardness Tester (AC220V/50 Hz:110 V/60 Hz, 0.01 mm of micrometer-resolution, and max. 95 mm horizontal distance).
Results and Discussion
The original curves of the DSC measurements of the CuAlBe alloy, at first cycled normally and then cycled with liquid nitrogen make-up during cooling, both taken at 5 °C/min of heating/cooling rate were given in Fig. 1 a and c. Also, these curves were given with their peak analyses on the time-axis in Fig. 1 b and d. In Fig. 1a, there can only be seen a down endothermic peak of the onward martensitic transformation from martensite to austenite (M → A) phase. As for the counter peak of the backward A → M transformation, this A → M peak was uncovered by lowering the temperature of the alloy sample in the DSC instrument below room temperature via using some liquid nitrogen make-up.
The characteristic martensitic peak start (onset) and finish (endset) temperatures of As, Af, Ms, and Mf were determined from the peak analyses of these two curves and given in Table 1. Here, the M → A transformation peak start and finish temperatures (As and Af) were determined from the peak analysis given in Fig. 1b because when this first curve was started to run, at the beginning of the DSC test, the as-quenched CuAlBe alloy sample was fully in martensite phase. Therefore, the alloy exhibited a complete M → A transformation in the first cycle. But then, in the second cycle, it exhibited a more shallow M → A peak on heating because the martensite phase was just partially formed in the alloy at the end of the first cycle, because the temperature was not lowered below the martensite finish (Mf) temperature by DSC instrument, because Mf (= 19.78 °C, as seen in Table 1) is a bit lower than room temperature. This is a small kind of kinetic stop because Mf is a bit lower than room temperature i.e. this DSC run (as seen in Fig. 1a) ended at room temperature and did not let the martensite be formed fully by cooling the alloy sample more until it reaches the Mf temperature. After all, these DSC results display a reversible martensitic transformation (with intermediate transformation temperatures, seen in Table 1) which indicates the presence of the SME property in the produced 77.05Cu-22.02Al-0.93Be (at.%) alloy in this work. The transformation temperatures for a previously studied Cu-22.7Al-3.1Be (at.%) alloy  were reported in between − 50 and − 15 °C (Ms = −37 °C) obtained 5–10 °C/min due to a higher Be content use than that (0.93Be at.%) used in this work. In another work , the temperature was found in a range of 124 °C and 175 °C (obtained at 10 °C/min of heating/cooling rate) for a Cu-14.82Al-0.4Be wt% (or 69.2Cu-28.49Al-2.3Be at.%) alloy due to the higher (6.47 at.% more) percentage of Al content (as a recompense of decrement in Cu percentage) that increased the transformation temperatures at least 100 °C. In another study  of a commercial Cu–11.8Al–0.47Be wt.% (or 72.86Cu-24.25Al-2.89Be at.%) alloy, the transformation temperatures obtained at 5 °C/min were found to be lower than 2.2 °C (= Af). So, it is clearly understood that Be reduces the transformation temperatures and contrarily Al increases them.
The other important kinetic parameters such as the hysteresis gap (As‑Mf), equilibrium temperature (To), and the enthalpy (ΔHM↔A) and entropy (ΔSM↔A) changes were determined and also listed in Table 1.
The values of entropy changes (∆SA↔M) caused by these both way transformations that are given in Table 1 were calculated by ∆SA↔M = ∆HA↔M /T0 relation [7, 9]. Here, the equilibrium temperature (To) is the temperature where both of the chemical Gibbs free energies of austenite and martensite phases are equalized. The value of T0 temperature was calculated by using T0 = (Af + Ms)/2 relation [9, 11].
The DTA cycle of the alloy was run at a heating/cooling rate of 25 °C/min and the result is given in Fig. 2. On this DTA curve, beginning from the start of heating on the far left, the CuAlBe alloy exhibited multiple phase transitions as β1(DO3) → α + γ2 decomposition (above 400 °C) → eutectoid recomposition (at ~ 513–550 °C) → B2(ordered) → A2(disordered), which is common in Cu-Al-(Be)-based SMAs [8, 9, 12,13,14,15]. The β1(DO3) superlattice of the austenite phase becomes unstable when the temperature of alloy approaches 400 °C and after this temperature, precipitations occur. By the end of the eutectoid reaction, these precipitations resolve again and the ordered cubic β(B2) phase occurs and then changes into the disordered β(A2) phase at higher temperatures.
By the EDS test made at room temperature, the chemical composition of the CuAlBe alloy was determined as 77.05Cu-22.02Al-0.93Be (at.%) which is used to calculate the average valence electron concentration per atom (e/a) value of the CuAlBe alloy, too. To know the e/a value of a shape memory alloy enables us to make an assumption, a theoretical estimation, on the existence of shape memory effect property and a comparison of the martensite phase types that formed in the alloy. The e/a ratio is a determining parameter for alloys to have shape memory effect property and for the vibrational entropy change (ΔS) of the average periodic non-diffusional atomic displacements in the crystal lattice that occur during first-order martensitic phase transformations [15, 16]. To show a shape memory effect, generally, the e/a values of Cu-based alloys should be in the interval between ~ 1.45 and ~ 1.49 . If the e/a value is lower than 1.45, then the monoclinical M18R (β1′ or β3′) martensite gains dominancy over the hexagonal 2H (γ1′ or γ3′) martensite in the alloy matrix, then both phases become nearly equal-volumetric in the mid of this interval, and then for the values above 1.49, the dominancy of 2H structure starts [15, 18]. In this work, the e/a value of the CuAlBe alloy was calculated by using e/a = ∑fivi formula [1, 15] which was found as 1.4497. Here, fi refers to the atomic fractions of alloying elements (by the EDS result), and vi is the valence electron numbers of these alloying metals. The theoretical estimation based on the e/a value that was mentioned above can be made also for the CuAlBe alloy and the experimental results of the microstructural tests given below are to support this estimation.
The XRD result for as-quenched CuAlBe alloy (i.e. it is in the martensite phase) is shown in Fig. 3. This X-ray diffraction pattern of CuAlBe alloy indicates that the alloy is a single-like crystal; i.e. it has a low degree of polycrystallinity, amorphism, or defects. The main and only high-intensity sharp peak on this pattern refers to the (120) plane which indicates β1′ martensite phase and the other small ones are β1′(042) and α(200) peaks [9, 18]. The main peak shows the dominancy of the β1′ martensite, just as the theoretical estimation above made upon the e/a ratio value of the alloy. Additionally, the average crystallite size (D) parameter of the alloy was found to be 29.32 nm by using D = 0.9λ/B1/2cosθ formula of Debye-Scherrer [18, 19], where λ is the wavelength (= 0.15406 nm) of the X-ray CuKα radiation, B1/2 is the full width at half maximum (FWHM) value for the highest peak (= 0.289), and θ is the Bragg angle of this diffraction peak (here, 2θ = 40.569° is for the highest β1′(120) peak). Compared with the others [7, 15, 18] reported previously, this larger sub-micrometer D crystallite size value of the CuAlBe alloy means that the coherently X-ray scattering domain (spherical or ellipsoidal particle) size of β1′ martensite phase in the alloy is larger and it has a higher single-like crystallinity which is good for shape memory property.
Hardness is often inversely related to ductility, so the ductile metals or alloys typically have relatively low hardness. The Vickers microhardness values obtained from three different regions of the alloy’s surface are 221 HV, 213 HV, and 201 HV, and on average 211.66 ± 3.36 HV. These microhardness values that refer to the good ductility or softness feature of CuAlBe alloy were found slightly consonant with those little higher values of 324.5 ± 3 HV , 263.9 ± 6.5 HV , and 223.9 ± 5.9 HV  reported previously in the literature. This means that the CuAlBe alloy is a bit softer than those reported alloys. The presence of Cu α-phase that was evidenced by the above-given XRD result conduced to the high softness of the alloy . The reason for the microhardness values of CuAlBe alloy being smaller is also due to the effect of the atoms of minor Be content on the lattice of Cu-Al texture. These small Be atoms acting like flexibility enhancer ligaments (or bridge or rivet) that bounded in between the bigger copper and aluminum atoms unstrung or loosened the alloying bonds between Cu and Al atoms and reduced their dipole-like rigidity and stiffness. Therefore, Be increased the softness that may enable more flexibility which may be useful in some bulk and micro-/nano-scale thin-film shape memory alloy applications.
The optical micrographs of two different regions of CuAlBe alloy surface obtained by × 250 magnification ratio were given in Fig. 4. As seen on these micro-images, the thin V-type and needle-like β1′(18R) martensites [5, 8, 9, 23] that formed on the surface of the alloy can be seen clearly. The thick and collaterally formed γ1′(2H) martensites (e.g., in the darker grain region, seen on the top right-hand corner of the left image) and some dark spots of AlCu3 (β1) precipitations [9, 10, 23,24,25,26,27] are also visible in Fig. 4.
In this work, the ternary CuAlBe shape memory alloy with a novel composition was fabricated by arc melting method and then underwent various kinds of thermal (DSC, DTA), mechanical (Vickers microhardness), and structural (XRD, EDS, and optical microscopy) measurements. The intermediate martensitic transformation temperatures were observed indicating the existence of the SME property and the formation of the β1′ martensite phase. While the Al element increases the transformation temperatures, Be element contrarily reduces the transformation temperatures. Multiple phase transition of β1(DO3) → α + γ2 → eutectoid reaction → B2 → A2 was observed as the alloy’s high-temperature behavior which is a common thermal behavior of CuAlBe shape memory alloys. The structural XRD and optical microscopy tests confirmed the theoretical estimation upon e/a ratio calculation by showing the existence of the dominant β1′ martensite over γ1′ martensite. The coherently X-ray scattering large domain size (crystallite size) of β1′ martensite phase in the alloy points to the single-like crystallinity of the produced alloy. The microhardness values that were lower than some previously reported ones demonstrated the improved ductility and higher softness obtained by the addition of minor Be content. The small Be atoms acting as flexibility enhancer ligaments between the bigger Cu and Al atoms in the lattice reduce the hardness and increase the softness that can enable more flexibility to be useful in some bulk and micro-/nano-scale thin-film shape memory alloy applications operating at these moderate martensitic transformation temperatures of the produced CuAlBe alloy.
C.M. Wayman, K. Otsuka, Shape memory materials (Cambridge University Press, 1998), pp. 1–50
K. Otsuka, X. Ren, Recent developments in the research of shape memory alloys. Intermetallics 7, 511–528 (1999). https://doi.org/10.1016/S0966-9795(98)00070-3
Y. Furuya, Design and material evaluation of shape memory composites. Int. J. Fatigue 10(19), 724–330 (1997). https://doi.org/10.1177/1045389X9600700313
R. Dasgupta, A look into Cu-based shape memory alloys: present scenario and future prospects. J. Mater. Res. 29(16) (2014). https://doi.org/10.1557/jmr.2014.189
Ferreño et al., Thermal treatments and transformation behavior of Cu-Al-Be shape memory alloys. J. Alloys Compd. 577, 463–467 (2013). https://doi.org/10.1016/j.jallcom.2012.02.006
J. Lerner, et al., The effect of tin on the susceptibility of a Cu–Al alloy to cracking in mercury, Mater. Sci. Eng.: A Volume 345 (1–2): 357–358, (2003). https://doi.org/10.1016/S0921-5093(02)00098-9
C.A. Canbay, O. Karaduman, N. Ünlü, İ. Özkul, An exploratory research of calorimetric and structural shape memory effect characteristics of Cu–Al–Sn alloy, Physica B: Condensed Matter. 580, 411932 (2020). https://doi.org/10.1016/j.physb.2019.411932
S.M. Chentouf, M. Bouabdallah, H. Cheniti, A. Eberhardt, E. Patoor, A. Sari, Ageing study of Cu–Al–Be hypoeutectoid shape memory alloy. Mater. Charact. 61(11), 1187–1193 (2010). https://doi.org/10.1016/j.matchar.2010.07.009
C. Aksu Canbay, A. Aydoğdu, Thermal analysis of Cu-14.82 wt% Al-0.4 wt% Be shape memory alloy. J. Therm. Anal. Calorim. 113, 731–737 (2013). https://doi.org/10.1007/s10973-012-2792-6
S.N. Balo, M. Ceylan, M. Aksoy, Effects of deformation on the microstructure of a Cu–Al–Be shape memory alloy. Mater. Sci. Eng. A 311(1–2), 151–156 (2001). https://doi.org/10.1016/S0921-5093(01)00927-3
H. Tong, C. Wayman, Characteristic temperatures and other properties of thermoelastic martensites. Acta. Metall. 22, 887–896 (1974). https://doi.org/10.1016/0001-6160(74)90055-8
M.O. Prado, P.M. Decorte, F. Lovey, Martensitic transformation in Cu-Mn-Al alloys. Scr. Metall. Mater. 33(6), 877–883 (1995). https://doi.org/10.1016/0956-716X(95)00292-4
S.M. Chentouf, M. Bouabdallah, J.C. Gachon, E. Patoor, A. Sari, Microstructural and thermodynamic study of hypoeutectoidal Cu–Al–Ni shape memory alloys. J. Alloy. Comp. 470, 507–514 (2009). https://doi.org/10.1016/j.jallcom.2008.03.009
M.T. Ochoa-Lara, H. Flores-Zúñiga, D. Rios-Jara, G. Lara-Rodríguez, In situ X-ray study of order–disorder phase transitions in Cu–Al–Be melt spun ribbons. J. Mater. Sci. 41, 4755–4758 (2006). https://doi.org/10.1007/s10853-006-0026-7
C.A. Canbay, O. Karaduman, N. Ünlü, S.A. Baiz, İ. Özkul, Heat treatment and quenching media effects on the thermodynamical, thermoelastical and structural characteristics of a new Cu-based quaternary shape memory alloy. Composites Part B 174, 106940 (2019). https://doi.org/10.1016/j.compositesb.2019.106940
M. Ahlers, Phase stability of martensitic structures. Le J Phy IV 5(C8), C8-C71–C88-80 (1995). https://doi.org/10.1051/jp4:1995808
J. Pelegrina, M. Ahlers, The martensitic phases and their stability in Cu-Zn and Cu-Zn-Al alloys-I. The transformation between the high temperature β phase and the18R martensite. Acta Metall. Mater. 40, 3205–3211 (1992). https://doi.org/10.1016/0956-7151(92)90033-B
Canbay et al., Investigation of varied quenching media effects on the thermodynamical and structural features of a thermally aged CuAlFeMn HTSMA. Phys. B. Cond. Matt. 557, 117–125 (2019). https://doi.org/10.1016/j.physb.2019.01.011
A. Patterson, The scherrer formula for X-ray particle size determination. Phys. Rev. 56(10), 978–982 (1939). https://doi.org/10.1103/PhysRev.56.978
J.P. Oliveira, Z. Zeng, S. Berveiller, D. Bouscaud, F.M. Braz Fernandes, R.M. Miranda, N. Zhou, Laser welding of Cu-Al-Be shape memory alloys: microstructure and mechanical properties, materials & design. Vol. 148, 145–152 (2018). https://doi.org/10.1016/j.matdes.2018.03.066
J.P. Oliveira, B. Panton, Z. Zeng, T. Omori, Y. Zhou, R.M. Miranda, F.M. Braz Fernandes, Laser welded superelastic Cu–Al–Mn shape memory alloy wires. Mater. Des. 90, 122–128 (2016). https://doi.org/10.1016/j.matdes.2015.10.125
J.P. Oliveira, B. Crispim, Z. Zeng, T. Omori, F.M. Braz Fernandes, R.M. Miranda, Microstructure and mechanical properties of gas tungsten arc welded Cu-Al-Mn shape memory alloy rods. Journal of Materials Processing Technology, Volume 271, 93–100 (2019). https://doi.org/10.1016/j.jmatprotec.2019.03.020
A. Agrawal, S.K. Vajpai, Preparation of Cu–Al–Ni shape memory alloy strips by spray deposition-hot rolling route. Mater. Sci. Technol. 36(12), 1337–1348 (2020). https://doi.org/10.1080/02670836.2020.1781354
J.d.B. Simões, E.M.A. Pereira, J.J.d.M. Santiago, C.J.d. Araújo, Microstructure and thermal analyses of Cu87-xAl13Nbx high-temperature shape memory alloys. Mater. Res. 22(Suppl. 1), e20180849. Epub November 28, 2019 (2019). https://doi.org/10.1590/1980-5373-mr-2018-0849
S. Montecinos, S. Tognana, W. Salgueiro, Determination of the Young’s modulus in CuAlBe shape memory alloys with different microstructures by impulse excitation technique. Materials Science and Engineering: A, Vol. 676, 121–127 (2016). https://doi.org/10.1016/j.msea.2016.08.100
S. Tognana, S. Montecinos, W. Salgueiro, Influence of quenched-in vacancies on the elastic modulus and its dependence on the temperature in β CuAlBe shape memory alloys, Intermetallics. Vol. 111, 106485 (2019). https://doi.org/10.1016/j.intermet.2019.106485
R.K. Singh, S.M. Murigendrappa, S. Kattimani, Investigation on properties of shape memory alloy wire of Cu-Al-Be doped with zirconium. J. of Materi Eng and Perform (2020). https://doi.org/10.1007/s11665-020-05233-7
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Canbay, C.A., Karaduman, O., Ünlü, N. et al. Study on Basic Characteristics of CuAlBe Shape Memory Alloy. Braz J Phys (2020). https://doi.org/10.1007/s13538-020-00823-1
- Ductile CuAlBe SMA
- Differential calorimetry
- Vickers microhardness test