Effect of Cryogenic Treatment on the Phase Transformation Temperatures and Latent Heat of Ni₅₄Ti₄₆ Shape Memory Alloy ^†

Abstract

Deep cryogenic treatment is characterized by subjecting the material to slow and controlled cooling at temperatures of up to approximately −196 °C or 77 K. This treatment has attracted industry interest and has been used in recent years, with relative success, to improve the properties of metals, especially steels. In this article, we provide a preliminary study on the cryogenic soaking time effect of 6, 12, 24, and 36 h on the temperatures and latent transformation heat of the Ni₅₄Ti₄₆ alloy phase. By differential scanning calorimetry (DSC), the properties of samples subjected to different cryogenic soaking times at −196 °C were determined. The results indicated an increase in the phase transformation temperatures present in the alloy and a decrease in the latent heat of transformation in relation to the cryogenic soaking time. The reduction of 9% in the energy involved in the phase transformation was probably due to microstructural changes, and the increase of up to 14.8% in phase transformation temperatures was probably due to stress relief from the reduction of dislocations.

1. Introduction

Shape memory alloys (SMAs) have potential application in various areas of engineering and medicine [1]. In many of the proposed applications for SMAs, the limitations of the thermomechanical behavior of SMAs such as superelasticity (SE) or pseudoelasticity and shape memory effect (SME) may represent an impediment to the application of SMAs [2]. Therefore, after a period of many application propositions of SMAs, intense research was developed in order to seek a way to improve the thermomechanical properties of SMAs [3].

Cryogenic treatments have been used to improve the properties of tool steels since the 1970s. Yun [4] and Mariante [5] observed that the wear resistance of fast steel AISI M2 increased by 51% with cryogenic treatment. Molinari [6] reported that, in the case of AISI M2 steel, the increase in wear resistance could be attributed to increased hardness, and in the case of AISI H13 steel, the increase in wear resistance was correlated with increased toughness.

Equiatomic NiTi alloys were studied by a number of authors who validated the profound changes in properties by deep cryogenic treatment (DCT). Currently, orthopedic and orthodontic applications are the most common ways NiTi alloys are used [2]. This has attracted the attention of researchers working on enhancing alloy properties to improve the machining and clinical performance of orthodontic appliances using DCT [7].

Avantika Singh [8] evaluated the efficacy of cryogenic treatment at −196 °C in NiTi endodontic instruments, made by ProTaper™. His method consisted of the comparative analysis of four endodontic instruments, and he empirically took the cutting efficiency from the mass loss of 24 tooth samples after instrumentation for 240 s. When analyzing the influence of cryogenic treatment on rotary orthodontic files, he observed that cryogenically treated files had little significance in terms of wear resistance, but there was a significant improvement in microhardness and greater cutting efficiency.

Kim [9] also investigated the effect of cryogenic treatment on austenitic phase NiTi orthodontic files that were previously submerged in liquid nitrogen at −196 °C for different time intervals. A total of 80 samples of Ni₅₆Ti₄₄ cutting instruments were used, where half of the instruments were cryogenically treated and evaluated for composition, microhardness and cutting efficiency. An increase in wear resistance and a slight increase in microhardness were evident according to the soaking time; however, there was no improvement in cutting efficiency.

Authors Kim [9] and Avantika [8] reached similar conclusions for the effects of DCT on endodontic instruments when relating the mechanical properties of microhardness. However, the hardness of the different NiTi alloys presented many variations after DCT [7]. As reported by current studies, for other alloys of the Ni-Ti system in the equiatomic vicinity, a decrease in hardness is commonly observed in cryogenically treated samples in relation to the control group [7,10].

Vinothkumar [10] performed microstructural characterization on Ni₅₁Ti₄₉ alloys with SME after treatment at approximately −196 °C in liquid nitrogen. Sheet and cylindrical-shaped specimens were used and were analyzed for hardness and wear resistance. Their results concluded that soaking time produced significant changes in the SME properties, such as variations in transformation temperatures and hardness of the NiTi. A decrease in hardness was found in contrast to an increase in martensitic fraction after the deep cryogenic treatment of 6 and 24 h.

SMAs are sensitive to heating and cooling cycles with regard to microstructural changes. Similar to steels, these alloys can undergo changes in mechanical, thermal, electrical, or chemical properties when subjected to appropriate heat treatments. Maintaining martensitic transformation temperatures is influenced by martensitic thermal cycling, as demonstrated by Da Silva [11].

In the alloys of the Ni-Ti system, the R-phase can be induced by thermal treatments and formed by the appearance of Ni₄Ti₃ precipitates [12]. The main advantage of heat treatments in NiTi alloys with shape memory effect is the reduction in stiffness associated with the B19’ phase [13]. The cryogenic treatment immersion time promotes the decrease of austenite to martensite growth upon the martensitic phase transformation for a cryogenic-treated NiTi [14]. Martensitic transformation temperatures are strongly affected by the microstructural characteristics of the alloy and grain size [15,16]. This study aimed to evaluate the influence of deep cryogenic treatment on phase transformation temperatures as well as the latent transformation heat of the Ni₅₄Ti₄₆ wire shape memory alloy phase.

2. Methods and Materials

The material investigated was NiTi SMAs marketed in the form of an actuator wire SmartFlex150 by SAES Getters Group (Colorado Springs, CO, USA), with a diameter of 0.15 mm, Ni content of 54% by weight, and A_F > 90 °C. The typical data of some physics-mechanical properties and characteristics provided by the manufacturer are given in

Table 1. SmartFlex^TM150 actuator properties data from [17].

Product	Diameter [µm]	Max. Force [N]	Max. Deformation	Suggested Force [N]	Suggested Deformation
SmartFlex150	150	6.2	5.5%	2.7	<3.5%
Hysteresis test: 200 MPa, 1 °C/minute
Recoverable deformation			4.8%
As			86 °C
Af			94 °C
Ms			65 °C
Mf			57 °C
Fadiga: 3.5% 170 MPa 0.6 A
Cycle numbers			>105
Variance			0.17%

. Other data from the material investigated were presented by Fumagalli [17]. The Ni-Ti binary phase diagram is shown in Figure 1. One could notice that this system contains the stable NiTi, Ni₃Ti, and NiTi₂ phases, and the metastable phases Ni₄Ti₃ and Ni₃Ti₂ can also be found.

The methodology adopted was a comparative analysis of the properties of the treated and untreated material cryogenically. For this end, five wire specimens were prepared and divided into two groups of knowledge: cryogenically treated and cryogenically untreated. The cryogenic process was by direct immersion of the specimens in liquid nitrogen. The specimens were cooled to −196 °C without rate control by maintaining this temperature for soaking times of 6, 12, 24, and 36 h, respectively, and subsequent heating naturally to the room temperature of 24 °C.

The phase transformation temperatures and the latent heat of transformation before and after cryogenic treatment were obtained by differential scanning calorimetry (DSC), model DSC 8500, by PerkinElmer™ (Waltham, MA, USA). The tests were performed again by using DSC Q20, by TA Instruments™ (New Castle, DE, USA), equipment for the same temperature range from −20 °C to 100 °C, with heating and cooling rates of 10 °C/min. The phase transformation temperatures were determined by means of the tangent method, which was applied to the phase transformation peaks of the DSC curves, according to international standards ASTM F2004, F2005 (ASTM, 2005), and F2082 (ASTM, 2006). The thermal hysteresis ∆HT was calculated by using the difference between the peak heating temperatures A_P and the cooling M_P.

3. Results and Discussion

3.1. The Phase Transformation Temperatures

In the DSC curve, the present temperatures were observed, as well as the peak transformation temperatures, M_P and A_P, of the non-cryogenic treated (NCT) and cryogenically treated (CT) specimens. The temperature phase evolution, the hysteresis as a function of soaking time, and the results of the DSC analysis for untreated specimens are shown in Figure 2 In addition, the average values of the characteristic phase transformation temperatures during the inverse transformation (heating) and the direct transformation (cooling) of the cryogenics for the treated specimens are indicated in Figure 3.

The single peak visible during warm-up could be due to the transition of the alloy from phase B19’ to phase B2. In cooling, the two main peaks are associated with two distinct martensitic transformations: the first is related to the B2-R transitions and the second to the R–B19’ transitions, where B2, R, and B19’ represent the cubic (austenite), the rhombohedral, and monoclinic structures (martensite), respectively.

The intermediate phase R is common in NiTi-based SMAs for compositions close to the equiatomic. The presence of the R phase is more remarkable when the SMA is subjected to appropriate heat and/or thermomechanical treatments [1,18]. This intermediate phase originates from the nucleation and growth of fine Ni₄Ti₃ precipitates, which introduce residual stresses and promote heterogeneity in the matrix phase [19].

Figure 2 shows the R-phase clearly as a low-intensity peak in the cooling curve of the NCT sample. This may indicate the formation of compounds deposited with different Ni compositions. Similarly, in the analyses with the CT samples, it is possible to observe the peak of similar phases with no significant differences in intensity or variation of the onset temperature R_S, except for sample CT6. Soaking for 6 h showed variation in R_S, with a peak of lower intensity. The variation in R_S and the decline in peak intensity may reflect solubilization or the non-formation of intermediate phases and Ni precipitates for treatment at shorter soak times.

The analysis of the curve shown in Figure 3 indicates that the alloy is completely in the martensitic phase at room temperature (24 °C), from which a quasi-plastic behavior is expected, followed by the shape memory effect if heated above A_F. In addition, the DSC curves (Figure 3) present an increase in temperatures M_S and M_F of all CT specimens in relation to NCT specimen, with the most remarkable change in specimens treated for 6 h and 12 h. On the other hand, no change was observed in temperatures A_S and A_F. Additionally, no change was observed for the transformation temperature average values of the specimens CT24 and CT36, except for the temperature A_F of specimen CT6, which presented a reduction of 1.92%.

Analyzing Figure 3a, the temperatures of the samples treated with CT6, for 6 h, contrast with the temperatures of the untreated samples, NCT, which exhibited increases of 12.27% in M_S and 8.5% in M_F. When the same comparison was made for the material cryogenically treated for 12 h, CT12 (Figure 3b), the increase was higher for M_S and lower for M_F, resulting in 14.81% and 5%, respectively. The observed increase in M_S and M_F temperatures was mainly due to the cryogenic treatment, and its intensity was higher in the sample treated for 6 h, as indicated in the graph of Figure 4.

The change in phase transformation temperatures was also reported by other researchers, and it is associated with the stress relief from reducing the density of dislocations [18]. The major difference in the temperature was observed in the shortest immersion time. This was justified in part by the small wire diameter and changes that occurred in the CT treated-specimen microstructure, e.g., the stress relief from the reduction in the density of dislocations and changes in grain size [12,16].

Transformation temperatures are very sensitive to variations in Ni content and increase their value with a decrease in the Ni ratio of the matrix phase [20]. These decreases in Ni content occur as a result of the depletion of Ni from the matrix phase and the diffusion to further nucleation and growth of intermediate phases and precipitates. Indeed, the deep cryogenic treatment is a method that modifies the entire structure of the material and can be used to promote the appearance of precipitates and other metastable particles [9].

3.2. Evolution of Hysteresis and Latent Heat

The transformation hysteresis data are presented in Figure 5, and the latent transformation heat ΔHA and ΔHM of the treated and untreated specimens are shown in Figure 6. It is noted that the transformation hysteresis of cryogenically treated CT specimens presented lower values than those of the untreated NCT. There was a reduction in hysteresis inversely proportional to the soaking time. The difference was more noticeable in cryogenically treated specimens for 6 and 12 h, which had a reduction in thermal hysteresis of 8.96% and 5.46%, respectively.

Regarding the latent transformation heat ΔHA and ΔHM of CT specimens, there was a reduction in values, except for specimen CT36, which showed an increase in relation to NCT, the untreated sample. Therefore, the cryogenic treatment resulted in the reduction of thermal energy involved in the phase transformation phenomenon. Castilho [15] obtained similar results for different Ni₅₅Ti₄₅ and Ni₅₇Ti₄₃ alloys and found a decrease in the ranges of variation in thermal hysteresis after 24 h of treatment and little significance at times greater than 36 h. Furthermore, reaction at a lower heat was observed for the samples treated for a shorter time. This is due to an increase in the grain size, which was observed in specimens with up to 12 h of treatment.

The literature indicates that the changes observed in the latent heat of transformation occur due to the formation of Ni-rich precipitates. Ti₂Ni precipitates are secondary phases commonly found in the microstructure of Ni-Ti system alloy, and they are said to be undesirable phases as they cannot be fully dissolved in the matrix and may originate in Ti-rich brittle oxides. For instance, Cruz [21] evaluated the influence of DCT on the thermomechanical properties of Ni₄₈Ti₅₂ alloy. From the comparative study between treated and untreated samples, he inferred through scanning electron microscopy (SEM) and X-ray diffraction (XRD) the presence of the titanium precipitate phase, Ti₂Ni, in all treated samples, which was also accompanied by a reduction of latent heat and hysteresis of the alloy.

In addition, Gontijo [22] analyzed the effects of cryogenic treatment on the cyclic behavior of Ni₅₇Ti₄₃ alloy under thermomechanical loading. For the study, part of the samples were cryogenically treated for 12 h with immersion in liquid nitrogen, and part was kept untreated. When evaluating both groups of samples in tensile and cyclic tests, a greater reduction in the hysteresis area of the CT samples and a reduction of the transformation stresses was noticeable in the NCT samples. Further microstructural analyses performed with scanning electron microscopy (SEM) and energy dispersive spectrometry (EDS) revealed an increase in grain size and a decrease of the martensitic fraction. Additionally, the presence of pores and TiC inclusions in all samples were remarkable. TiC particles cause a plastic mismatch between the NiTi matrix, which generates residual stresses that may influence the phase change temperatures and hysteresis of the alloy [23].

4. Conclusions

The Ni₅₄Ti₄₆ shape memory alloy was subjected to cryogenic treatments by immersion in liquid nitrogen for 6, 12, 24, and 36 h. The results presented showed that the cryogenic treatment influenced the properties of the alloy. The temperatures at the beginning of martensitic transformation increased with the soaking time of −196 °C, with the intensity being higher in treatments for 6 h and 12 h, with increases of 12.2% and 14.81%, respectively. On the other hand, a certain temperature stability at the beginning and the end of the austenitic transformation was observed for the average values A_S, and A_F remained stable for the cryogenic treatments with soaking times above 12 h.

Regarding the latent heat of transformation of the phase, there was a decreasing trend in relation to cryogenic soaking time. The reduction was greater in the cryogenically treated material for 6 h. In this case, the reductions observed were 23% and 17% for ΔHM and ΔHA, respectively. The same behavior was observed in the transformation hysteresis, which had a reduction of approximately 9%.