1. Introduction
NiTi shape memory alloys (SMAs) have been widely utilized in different fields of aerospace and medicine due to their superior shape memory effect and higher superelasticity compared to other shape memory alloys [1]. It has been reported that the shape memory properties of NiTi alloys can be modified by adding ternary elements such as Al and Cu which are chemically similar to Ti and Ni, respectively [2]. The addition of Al to NiTi alloys is quite attractive for the remarkably improvement of room and high temperature strength [3]. Therefore, NiTiAl alloys are desirable candidate materials for high temperature structure applications. On the other hand, the addition of Cu to NiTi alloys reduces the sensitivity of the transformation temperatures to the chemical composition and to the thermal treatment time, diminishes the thermal hysteresis, reduces the mechanical hysteresis in the pseudoelastic regime and enhances the thermomechanical cycling stability [4].
To date, several methods have been developed to synthesize NiTi-based alloys such as conventional powder metallurgy, self-propagating high temperature synthesis (SHS), shock synthesis and mechanical alloying (MA) [5]. Among these methods, MA is a powder technique that allows production of homogeneous materials from blended elemental powder mixtures. In this method, melting is not essential and the products have nanostructural characteristics as a result MA has been applied to produce all material classes including metals, ceramics, and polymers [6]. In the field of metallic structures, a number of investigations were carried out to prepare a variety of stable and meta-stable phases, including supersaturated solid solutions, crystalline and quasi-crystalline intermediate phases, and amorphous alloys [7]. However, there are a few papers about the structural characterization of the nanocrystalline Ni(50−X)Ti50CuX alloys produced by MA process [8,9].
In this work, MA of the Ni(50−X)Ti50CuX (X = 5, 9 wt%) system and the structural features of the experimental outcomes were examined. Also, the phase transitions and fraction of crystalline phase after MA process at different circumstances were investigated.
2. Materials and Methods
2.1. Synthesis of Nanocrystalline Alloys
The powders of Ti (Merck, 98%), Ni (Scharlau, 99.8%) and Cu (Sigma Aldrich, 99.5%) were used as raw materials. The elemental powders with the nominal composition of Ni(50−X)Ti50CuX (X = 5, 9 wt%) were mechanically alloyed in a planetary ball mill under a high purity argon (99.998 vol%) atmosphere. The high purity argon was served to prevent extra phase formation such as TiNi during MA. Mechanical activation was carried out at room temperature in tempered chrome steel vials (vol. 125 ml) and balls (20 mm in diameter). The ball-topowder weight ratio and the rotational speed were 20:1 and 600 rpm, respectively. In all experiments the total powder mass was 6 gr. To prevent the agglomeration of powders during milling, 1 wt% of stearic acid was used as a process control agent (PCA). In addition, 45 min ball milling duration was followed by an interval of 15 min in order to avoid excessive temperature rise within the grinding vial. Details of process specifications and abbreviated name of products are given in Table 1.
2.2. Characterization of Nanocrystalline Alloys
Phase analysis of products was carried out by X-ray diffraction (Philips X-ray diffractometer (XRD), Cu-Kα radiation, 40 kV and 30 mA). For qualitative analysis, XRD graphs were recorded in the interval 10˚ ≤ 2θ ≤ 70˚ at scan speed of 1˚/min. “PANalytical X’Pert HighScore” software was used for the analysis of the diffraction patterns. The patterns were compared to standards compiled by the Joint Committee on Powder Diffraction and Standards (JCPDS), which involved card #001-1260 for Ni, #044-1294 for Ti, #004-0836 for Cu and #019-0850 for NiTi. Average crystallite size and lattice strain of the samples were determined using the XRD data according to the following equations [10]:
Table 1. Specifications of the synthesis process and abbreviated name of specimens.
(1)
(2)
where b (in radians), K, λ, D, E and θ are the structural broadening, shape coefficient (between 0.9 and 1.0), the wavelength of the X-ray used (0.154056 nm), crystallite size, lattice strain and the Bragg angle (˚), respectively. If we assume that a crystallite is a sphere of diameter D surrounded by a shell of grain boundary with thickness t, the volume fraction of grain boundary, f, is approximately [11]:
(3)
Values of f were calculated from this equation by substituting the experimental crystallite size obtained by XRD with D under the assumption of t = 1 nm.
Moreover, for the cubic structure, the lattice parameter a of a given plane with Miller indices (hkl) can be derived using the following equation [12]:
(4)
The (111) and (110) reflections were chosen for the lattice parameters calculation of Ni and B2-NiTiCu, respectively. In addition, volume of the cubic unit cell was determined by the following formula [12]:
(5)
The lattice parameter and unit cell volume deviations (in terms of %) were shown as [1]:
(6)
(7)
∆a and ∆V are defined as: ∆a = a − a0 and ∆V = V – V0, respectively. Where a0 is the standard value of lattice parameter which equals to 0.35175 nm for Ni and 0.29720 nm for B2-NiTiCu and a is the lattice parameter of the specimens. Also, V0 is the standard value of unit cell volume which equals to 0.04352 nm3 for Ni and 0.02625 nm3 for B2-NiTiCu and V is the lattice parameter of the samples.
The fraction of crystalline phase (crystallinity) was determined from the XRD data using the following equation [13]:
(8)
where Xc, K and B are the fraction of crystalline phase, a constant found equal to 0.24 and FWHM (˚) of selected reflection peaks, respectively.
3. Results and Discussion
3.1. Phase Evolution
Figure 1 shows the XRD patterns of the Ni45Ti50Cu5 mixed powders as a function of MA time. For comparison, the XRD patterns of the standard samples including #001-1260 for Ni, #044-1294 for Ti and #004-0836 for Cu are also shown in the same figures. At early stage of MA (milled for 5 min), the phase compositions were Ni, Ti and Cu. After 300 min of milling, broadening of the peaks and a significant decrease in the line intensity of Ni and Ti were dominant. Besides, all the peaks corresponding to Cu were vanished and B2-NiTiCu as a new phase was emerged. With further MA up to 600 min, the Ni and Ti diffraction lines were disappeared completely, and only broadened fundamental B2-NiTiCu peaks were observed.
Figure 2 displays the XRD profiles of the Ni41Ti50Cu9 mixed powders after different milling times. Similar to previous figure, the XRD patterns of the standard samples including #001-1260 for Ni, #044-1294 for Ti and