INTRODUCTION

Shape memory alloys (SMAs) of the Ti–Ni–Fe system have usually been used for the thermomechanical connection of pipelines [1, 2]. However, couplings made of these alloys must be mandrel-pressed, stored, and installed at cryogenic temperatures. Russian researchers have therefore suggested using additional alloying of titanium nickelide with niobium to expand the martensitic hysteresis of SMAs [3–5]. A great number of papers on this topic are given in [6]. We investigated Ti–Ni–Nb SMAs in the cast state, and, later, in the pressed-formed condition [7–10]. The effect of a number of factors on the properties of Ti–Ni–Nb SMAs was studied in detail in [5, 6, 11, 12]. However, a comprehensive study of the effect of various prestraining conditions on the shape memory effect (SME) in the 45Ti–45Ni–10Nb (at %) alloy was not performed. Therefore, the aim of this work is to study the effect of the temperature and rate of applying preliminary straining on the phase transformation parameters and thermomechanical characteristics of the 45Ti–45Ni–10Nb (at %) alloy in the press-formed state.

EXPERIMENTAL

We studied the 45Ti–45Ni–10Nb (at %) SMA of batch no. 193-11-p delivered as a pressed rod 25 mm in diameter. Samples and the metallographic polished section to be investigated were prepared. Samples for X-ray structural studies are described in [9], and for the study of thermomechanical characteristics, in [7]. We annealed the samples under the following conditions: heating in a vacuum furnace to a temperature of 850°C, holding for 4 h, and furnace cooling. The samples were then subjected to tensile tests (for prestraining) under various temperature, rate, and deformation conditions.

The phase composition, crystal lattice and substructural parameters, and characteristics of martensitic transformations (MTs) in the alloy samples in the initial state and after annealing were determined using an X-ray diffractometer in Cu Kα radiation at a power of 18 kW, as described in [3]. Tensile tests of the samples were carried out on a UTS-100K testing machine to find the thermomechanical characteristics of the alloy and to build stress–strain (σ–ε) diagrams. After deformation, the investigated samples were placed into a thermochamber of an R1084 unit to heat the samples to a temperature of Т = 100°С at a heating rate of 4 K/min. Then, the samples were cooled to Т = ‒170°С. Heating made them shorter and resulted in the shape memory effect. Upon cooling, no reversible shape memory effect was observed. The shape recovery diagrams of the alloy samples were used to determine characteristic temperatures \(A_{{{\text{s SME}}}}^{{\text{b}}}\) and \(A_{{{\text{f SME}}}}^{{\text{e}}}\) at the beginning and at the end of the entire shape recovery range, respectively, where SME manifests itself. The tangential method was used to find characteristic temperatures Аs SME and Аf SME, which characterize the main shape recovery in the \(A_{{{\text{s SME}}}}^{{\text{b}}}\)−\(A_{{{\text{f SME}}}}^{{\text{e}}}\) temperature range. The degree of shape recovery ηSME during SME was found from

$${{\eta }_{{{\text{SME}}}}} = \frac{{{{\varepsilon }_{{{\text{SME}}}}}}}{{{{\varepsilon }_{{\text{r}}}}}},$$

where εSME is the thermally reversible deformation during SME and εr is the residual deformation after unloading [14].

RESULTS AND DISCUSSION

1.MT in the course of development of SME during heating in the 45Ti–45Ni–10Nb alloy after its preliminary deformation by tension. The X-ray diffraction (XRD) analysis was used to investigate reverse MT in the alloy samples during heating-induced SME after preliminary deformation by tension under various conditions (varying the strain ε0 preliminary to SME-initiating holding at constant temperature Td and deformation rate \(\dot {\varepsilon }\), or changing Td at fixed ε0 and \({\dot {\varepsilon }}{\text{,}}\) or changing \({\dot {\varepsilon }}\) at fixed ε0 and Td).

Figure 1 shows an example of XRD patterns taken at different total prestraining deformations ε0 = 6% and ε0 = 30%. The study allowed us to build martensitic curves to determine the temperatures of the reverse MT (\(A{{'}_{{{\text{s SME}}}}}\), \(A{{'}_{{{\text{f SME}}}}}\) are the temperatures of the start and finish of the reverse MT during SME). The measurement error was ±5 K.

Fig. 1.
figure 1

X-ray diffraction patterns taken at different temperatures for the press-formed 45Ti–45Ni–10Nb alloy after tension ε0 = (а) 6 and (b) 30% (at Тd = –60…–70°С, \(\dot {\varepsilon }\) ≈ 1.2 × 10–3 s–1).

Full size image

Table 1 suggests that 100% B19' martensite is formed upon deformation of the alloy from 6 to 30% by tension at –60…–70°C. Subsequent heating in the temperature range from –60°С to 100°С results in the SME and the reverse MT of the TiNi phase. The MT in samples deformed to 6, 11, and 15% occurs completely and it develops according to the scheme В19' → В2. In samples deformed to 30%, the transformation is incomplete and it develops according to the scheme B19'→ (B2 + B19'). The remarkable thing is that heating in this case was carried out only up to a temperature of 100°C, because of the limited capabilities of the equipment employed.

Table 1. Schemes and the temperatures of the onset and finish of the reverse MT in the annealed samples of the press-formed 45Ti–45Ni–10Nb (at %) alloy during SME after tension (preceding the heating that initiates SME) performed under various conditions
Full size table

Martensite B19' is also formed upon the deformation of the alloy samples by tension at different temperatures (at a constant prestraining deformation equal to 11% and the deformation rate equal to 1.2 × 10–3 s–1). As the deformation temperature increases from –60…–70 to 24°С, the amount of the martensite decreases. The fraction of the martensite formed upon deformation at 24°C is 55%.

Subsequent heating from –60°С to 100°С causes complete MT according to the scheme В19' → В2 in the samples deformed at –60…–70 and 0…‒5°С. The transformation in the samples deformed at a temperature of 24°С develops according to the scheme (В19' + В2) → (В2 + В19') and it is not completed.

The deformation of the samples by tension at different deformation rates and a constant prestraining deformation that is 11% and a temperature of ‒60…–70°С causes the formation of 100% В19' martensite. Its amount remains the same even after the deformation rate is increased from 1.2 × 10–3 to 1.2 × 10–1 s–1. The reverse B19' → B2 MT is observed in all the samples upon heating from –60°C to 100°С.

Table 1 suggests that the temperatures of the reverse MT increase when the value of the prestraining deformation ε0 increases from 6 to 30%.

No effect of the deformation temperature in the range from –60…–70 to 0…–5°С on the temperature of the reverse MT was observed. The temperature range of the reverse MT \(\left| {A{{'}_{{{\text{s SME}}}}} - A{{'}_{{{\text{f SME}}}}}} \right|\) expands when the second deformation temperature range is from 0…–5 to 24°С. In this case, the temperatures of the reverse MT during SME significantly increase. Here, the shift of temperature \(A{{'}_{{{\text{f SME}}}}}\) is pronounced. This is the result of high-temperature SME associated with the stabilization of plastically deformed martensite [15].

Analysis of the effect of the prestraining deformation rate on the temperatures of the reverse MT showed that an increase in the deformation rate from 1.2 × 10–3 to 1.2 × 10–1 s–1 has almost no effect on the temperatures of the reverse MT (except temperature \(A{{'}_{{{\text{s SME}}}}}\)) and their range \(\left| {A{{'}_{{{\text{s SME}}}}} - A{{'}_{{{\text{f SME}}}}}} \right|\).

2.Phase composition and the structure of the 45Ti–45Ni–10Nb alloy in the course of SME after prestraining deformation by tension performed under various conditions. The XRD analysis of the phase composition of the alloy showed three phases (Table 2):

Table 2. Phase composition at 30°С in the annealed samples of the press-formed 45Ti–45Ni–10Nb alloy after SME induced by prestraining tension under various conditions
Full size table

—the main TiNi phase is present in two states (Table 2), namely, in the form of В2 austenite with an ordered bcc structure and B19' martensite with a distorted monoclinic orthorhombic crystal lattice (except for the prestraining deformation ε0 = 6%);

—bcc Nb phase; and

—a small amount of the fcc Ti2Ni phase.

Table 3 lists the parameters of the substructure and the TiNi (B2)-phase crystal lattice in the alloy.

Table 3. TiNi (B2) substructural and lattice parameters of the annealed samples of the press-formed 45Ti–45Ni–10Nb alloy after SME induced by prestraining tension under various conditions
Full size table

The lattice parameter of the TiNi (В2) phase after prestraining deformation of the 45Ti–45Ni–10Nb alloy to ε0 = 15–30% (at Td = –60…–70°С and \(\dot {\varepsilon }\) ≈ 1.2 × 10–3 s–1), as well as at a prestraining deformation temperature of 24°C (ε0 = 11%, \(\dot {\varepsilon }\) ≈ 1.2 × 10–3 s–1; see Table 3) is significantly different from that given in a handbook [16] (а = 3.015 Å for a two-component Ti‒Ni alloy).

This is because the TiNi phase under the above-mentioned prestraining deformation conditions exists in the intermediate two-phase (В2 + В19') state (see Table 2). The amount of the B19' phase is more than 5% (the error of the determination of the phase content in the alloy by the method of semiquantitative analysis is within 5%). The crystal lattice of the В19' phase is matched with the crystal lattice of the В2 phase and the atoms at the interphase boundary obey the crystalline order characteristic of both phases. This good matching between the two lattices requires a certain elastic deformation in the location of joining, which generates an elastic field of coherent stresses. Stresses change the lattice parameter [17].

The minimum number of structural defects (dislocations) characterizes the dislocation substructure of the alloys, which was formed after SME resulted from preliminary tensile deformation ε0 in the range from 6 to 11% (at Td = –60…–70°С, \({\dot {\varepsilon }}\) ≈ 1.2 × 10–3 s–1), at temperatures Td in the range from –60…–70°C to 0…–5°C (at ε0 = 11%, \(\dot {\varepsilon }\) ≈ 1.2 × 10–3 s–1), and deformation rates \(\dot {\varepsilon }\) from 1.2 × 10–3 to 1.2 × 10–1 s–1 (at ε0 = 11%, Тd = –60…–70°С). Correspondingly, the alloy contains fewer obstacles for MT propagation (Table 3).

The MT in this case takes place at lower temperatures and its rate is higher; that is, its range is narrower (Table 1).

3. Thermomechanical characteristics of the 45Ti–45Ni–10Nb SMA at development of SME after prestraining deformation by tension performed under various conditions.Figure 2 shows typical shape recovery diagrams taken during SME exhibited by the samples preliminarily deformed at different rates.

Fig. 2.
figure 2

Shape recovery curves of the 45Ti–45Ni–10Nb alloy during SME after prestraining tension by ε0 = 11% at Т = ‒40…‒45°C and different (prestraining) deformation rates \(\dot {\varepsilon }\): (1) 1.2 × 10–3, (2) 1.2 × 10–2, and (3) 1.2 × 10–1 s–1.

Full size image

Average thermomechanical characteristics resulted from the processing of the diagrams (considering the instrumental error) are listed in Table 4.

Table 4.   Thermomechanical characteristics of the press-formed and annealed 45Ti–45Ni–10Nb alloy during SME after preliminary tension performed under various conditions
Full size table

The statistical processing and the correlation analysis of the results [18–20] listed in Table 4 have revealed a significant effect of the value and temperature of the prestraining deformation preceding the initiation of SME on every thermomechanical characteristic of the alloys.

The temperatures of the beginning of shape recovery \(A_{{{\text{s SME}}}}^{{\text{b}}}\) = 42°С and Аs SME = 74°С, which are above room temperature, in the press-formed 45Ti–45Ni–10Nb alloy can be achieved at a preliminary (to SME) prestraining tensile deformation of ε0 = 15% at a rate \(\dot {\varepsilon }\) ≈ 1.2 × 10–3 s–1 and Тd = –60…–70°С. The values of the SME characteristics are maximum (εSME = 5.7%, ηSME = 0.5).

Table 4 suggests that the temperatures of the beginning of shape recovery \(A_{{{\text{s SME}}}}^{{\text{b}}}\) = 42°С and Аs SME = 52°С, which are above room temperature, and the range |Аs SME – Аf SME| = 42 K at the deformation temperature Тd = 24°С can be achieved in the alloy when temperature Тd of preliminary tension (deformation rate \(\dot {\varepsilon }\) ≈ 1.2 × 10–3 s–1 and total deformation ε0 = 11%) is increased from –60…–70 to 24°C. However, the SME characteristics significantly decrease: εSME decreases from 6.1 to 1.1% and ηSME, from 0.68 to 0.16. The SME characteristics (εSME = 6.1%, ηSME = 0.68) of this alloy can be maximum at the deformation temperature Td = 0…–5°C, however, the temperature of the beginning of the shape recovery is minimal (\(A_{{{\text{s SME}}}}^{{\text{b}}}\) = 15°C).

Analysis of data listed in Table 4 showed that an increase in the deformation rate from 1.2 × 10–3 to 1.2 × 10–1 s–1 does not change temperatures \(A_{{{\text{f SME}}}}^{{\text{e}}}\), Аs SME, and Аf SME, as well as εSME, ηSME, whereas it decreases \(A_{{{\text{s SME}}}}^{{\text{b}}}\) from 32 to 24°C and |Аs SME – Аf SME| from 13 to 9°C.

As a result, the study of the press-formed 45Ti–45Ni–10Nb alloy described in Sec. 3 showed that the alloy can be used for manufacturing the working element of the fire-protection device that we develop.

CONCLUSIONS

(1) Our study has shown that the best thermomechanical characteristics of the 45Ti–45Ni–10Nb (at %) alloy samples can be achieved by deforming the samples in the range 11–15% at temperatures of ‒60…–70°C and a deformation rate of 1.2 × 10–3 s–1. These deformation conditions can provide an optimum combination of the shape memory characteristics (εSME = 4.4–5.7%, ηSME = 0.5–0.53) and the temperatures of the shape recovery onset (\(A_{{{\text{s SME}}}}^{{\text{b}}}\) = 23–42°С, Аs SME = 65–74°С).

(2) We used the results of the study to develop a fire-protection device for nuclear power facilities.