Structure and Properties of TiNi Shape Memory Alloy after Quasi-Continuous Equal-Channel Angular Pressing in Various Aged States

Abstract

The effect of quasi-continuous (QC) equal-channel angular pressing (ECAP) in various pre-aged states on the structure formation and mechanical and functional properties of a hyper-equiatomic titanium nickelide (TiNi) shape memory alloy is studied. QC ECAP with a channel intersection angle of 110° is carried out at a temperature of 450 °C after aging for 1 and 5 h for three passes. To investigate the obtained structure and properties, the following research methods are applied: transmission electron microscopy, XRD analysis, calorimetric study, tension and hardness tests, and a special technique for the determination of functional properties. QC ECAP allows for the considerable refinement of structural elements and results in obtaining a mixed fine-grade structure, with structural elements of average sizes of 92 nm after pre-aging for 1 h and 115 nm after pre-aging for 5 h. Pre-aging for 5 h before QC ECAP, in combination with QC ECAP and post-deformation aging at 430 °C for 1 h, provides the best combination of mechanical and functional properties: a dislocation yield stress of 1410 MPa, ultimate tensile strength of 1562 MPa, and total recoverable strain of 11.6%. These values are comparable with the best results obtained for titanium nickelide and expand opportunities for the application of smart shape memory devices.

1. Introduction

One of the urgent problems of modern materials science is the search for new, and the improvement of existing, materials with unique physical, chemical, and mechanical properties. Among a wide class of smart materials, shape memory alloys (SMAs) are of great importance. The most promising type of SMA are alloys based on titanium nickelide (TiNi). TiNi SMAs with a Ni content of more than 50.5 at.% are widely applied for medical application due to the required temperatures of shape recovery near human body temperatures [1,2,3,4]. The development of the implementation of TiNi SMAs is associated with the improvement of mechanical and functional properties. The properties of these alloys strongly depend on a structural-phase state, which is formed after the usage of a particular thermomechanical treatment (TMT) [5,6,7,8]. In [6], it was established that the deformation of TiNi SMA in the temperature range of 300–600 °C contributes to the formation of a dynamically polygonized structure, which is most favorable for obtaining an ultrafine-grained structure. Additionally, for hyper-equiatomic TiNi alloys, the development of aging processes in the temperature range between 300 and 500 °C, consisting in the precipitation of Ti₃Ni₄ particles and the depletion of the matrix with nickel, has a great impact on the structure, the kinetics of martensitic transformation, and the combination of properties. The application of dynamic and static aging provides the ability to precisely change operational characteristics, especially the temperature range of shape recovery within a wide limit [8,9,10]. The precipitation behaviors of the Ti₃Ni₄ phase are largely dependent on chemical composition, aging temperatures/times, and the applied stress state during the aging treatment [10,11,12]. Additionally, the particular structural state of an alloy before aging strongly affects Ti₃Ni₄ particle size and morphology, as well as the temperature ranges of martensitic transformation and the obtained properties [13,14]. The application of the aging process before mechanical treatment significantly affects the deformation behavior and structure formation of TiNi SMA. In [15], it was found that aging at 773 K before cold working to form a coherent fine precipitate of Ti₃Ni₄, and subsequent cold working and post-deformation annealing at 673 K to form a nanocrystalline grain containing Ti₃Ni₄ particles, led to precipitation strengthening and grain-refinement strengthening, providing the best cycling stability and a drastic improvement of the superelasticity and the elastocaloric effect of a Ti-50.8 at.% Ni alloy.

However, in some cases, the application of pre-aging before the accumulation of higher strain may lead to the earlier distraction of the sample, especially during processes of severe plastic deformation (SPD). One of the most applied methods of SPD that allows the improvement of the mechanical and functional properties of TiNi SMA bulk samples is equal-channel angular pressing (ECAP) [16,17,18]. The effect of pre-aging treatment on the ECAP process and structure formation in TiNi hyper-equiatomic alloys was studied in [19,20]. It was found that Ti₃Ni₄ particles may dissolve into the matrix during the ECAP process. The degree of redissolution of particles is controlled by plastic deformation strain and the initial size of precipitates. The critical size for redissolution was defined to be from 20 to 38 nm after one ECAP pass.

While it was explained how static aging before ECAP affects the morphology, size, and distribution of Ti₃Ni₄ particles, the effect of pre-aging treatment before ECAP on the mechanical and functional properties of TiNi alloys is unclear. Additionally, during the traditional ECAP process, the precipitation or dissolution of Ti₃Ni₄ particles may occur during intermediate heating between passes. It was shown in [17] that the application of quasi-continuous (QC) ECAP can considerably refine the structure in equiatomic TiNi alloys and significantly improve their mechanical and functional properties. Therefore, it is necessary to study the influence of QC ECAP without intermediate heating between passes on the evolution of the Ti₃Ni₄ particle precipitation process. The present work aims to carry out a comparative study of the effect of static aging before ECAP, in combination with the dynamic aging process or redissolution of Ti₃Ni₄-phase particles during QC ECAP and post-deformation static aging, on the structure formation and mechanical and functional properties of a hyper-equiatomic TiNi alloy. An ECAP temperature of 450 °C and three passes were chosen as the most favorable conditions for the formation of optimum structure homogeneity in combination with the most intensive development of dynamic polygonization and aging processes [21,22].

2. Materials and Methods

For the QC ECAP procedure, hot rolled and polished rods of hyper-equiatomic TiNi SMA (Ti-50.7 at.% Ni) with a diameter of 20 mm were used. The total content of impurities in the rods was less than 0.25 wt.%, and the oxygen content did not exceed 0.1 wt.%. Samples for ECAP with a length of 85 mm were cut from supplied rods on the disk cut-off machine. Before QC ECAP, annealing at 750 °C for 30 min with quenching in water was carried out for all samples. The quenched state of TiNi samples was also used as a reference treatment (RT). After RT, the B2-austenite grain size in studied TiNi alloy rods was approximately 25–35 microns, and particles of the Ti₃Ni₄ phase were not indicated. The finishing temperature of reverse martensitic transformation, investigated using differential scanning calorimetry, was 8 °C.

QC ECAP was performed at a temperature of 450 °C with the following parameters of an ECAP matrix: a channel intersection angle, φ, of 110°, an outer angle, ψ, of 25°, and a pressing speed of 1.5 mm/s. Before ECAP, samples were annealed at 450 °C for 1 h (RT + 450 °C/1 h) and 5 h (RT + 450 °C/5 h) in order to achieve a different aged state, and immediately transferred to the container, which had been preliminarily heated to the same temperature of 450 °C and pressed in 3 passes in QC mode (RT + 450 °C/1 h +QC ECAP; RT + 450 °C/5 h + QC ECAP) [17].

All samples for the research procedure were obtained from the workpiece after QC ECAP using an electrical discharge machine. For the study of the effect of static post-deformation aging on the structure and properties after QC ECAP, post-deformation aging (PDA) at a temperature of 430 °C for 1 h followed by water quenching was performed (RT + 450 °C/1 h + QCECAP + 430 °C/1 h; RT + 450 °C/5 h + QC ECAP + 430 °C/1 h).

Phase composition was analyzed via X-ray diffraction (XRD) analysis using a DRON-3 X-ray diffractometer in Cu_Kα radiation in the 2ϴ angle range from 40 to 45°. The microstructure was examined via transmission electron microscopy (TEM) using a JEM-2100 transmission electron microscope at an accelerating voltage of 200 kV. Thin foils for TEM analysis were prepared through mechanical polishing at the first stage, followed by electrolytic polishing in a HCLO₄ + CH₃COOH acid solution. The temperature range of martensitic transformation was investigated via differential scanning calorimetry (DSC) using a Mettler-Toledo DSC 3+ calorimeter in the temperature range from −100 °C to 100 °C at a heating–cooling rate of 10 K/min in accordance with the ASTM standard F2004-16. Tensile tests were conducted using a INSTRON 3382 universal tensile machine at a deformation rate of 2 mm/min. Dogbone samples with the following dimensions were applied: an overall size of 1.0 mm × 14.0 mm × 70.0 mm and a gauge section of 1.0 mm × 7.0 mm × 20.0 mm. The instrumental error limits of the reported values were as follows: ±15 MPa for σ and ±1.5% for δ. Vickers hardness measurements were carried out using a LECOM 400-A tester under a load of 1 N and a dwell time of 10 s. No less than 10 indentations were carried out for each sample.

Functional properties were estimated via a thermomechanical method using the bending mode for strain induction [23]. Samples were bent around a mandrel at a temperature of 5 °C to induce a strain (Ɛ_t) of about 12.0%, followed by unloading and heating above the finishing temperature of reverse MT, A_s, for complete shape recovery followed by cooling to room temperature. The total recovery strain ($Ɛ$_rt), which includes the strain recovered due to the effect of superelasticity $(Ɛ$_se) right after unloading and the strain recovered after heating due to the shape memory effect $(Ɛ$_sm), and the residual strain ($Ɛ$_f) were determined [24]. The total induced strain (Ɛ_t) was determined based on the following relationship: Ɛ = d/(D + d), where d is the diameter of the sample and D is the diameter of the mandrel. The absolute error limit for $Ɛ$_r is ±0.3%.

3. Results and Discussion

3.1. X-ray Diffraction Analysis

The X-ray diffraction profiles in the vicinity of a major {110}_B2 peak after the studied regimes of heat and thermomechanical treatments included a RT and pre-aging in combination with QC ECAP and PDA, as presented in Figure 1.

After RT, only one narrow {110}_B2 prime peak, corresponding to the B2-austenite phase, was observed in the diffractogram. A half-height peak width, B₁₁₀, of 0.18 2Θ deg. is typical for recrystallized austenite with low dislocation density. A slight asymmetry peak was observed due to unresolved 110_α1–110_α2 doublet reflexes. Some traces of TiC phase were also present (see (200)_TiC line, indicated at 42 deg.). B19′-martensite lines were not detected, since the XRD study was carried out at room temperature (about 22 °C).

The aging of a TiNi alloy in the RT state for 1 h at 450 °C leads to the broadening of the B2-austenite line up to 0.30 deg., presumably due to the precipitation of Ti₃Ni₄-phase coherent particles and the appearance of the R-phase at room temperature as a result of the initiating effect of Ti₃Ni₄ precipitates on B2→R transformation. The latter is directly observable in the change in the profile shape because the abnormally large width of the lower part of the peak is formed by the (330)R–($3\overline{3}0$)_R doublet. An increase in the aging time from 1 to 5 h results in the appearance of two distinct peaks corresponding to R-phase (330)_R and ($3\overline{3}0$)_R lines, which indicates the additional initiating effect on the B2→R transformation because of the increase in the amount of the Ti₃Ni₄ phase. Note that the conclusion about the higher amount of the R-phase after aging for 5 h correlates well with the findings of DSC studies (Section 3.3).

The first thing that attracts attention when looking at the diffractograms after QC ECAP is a pronounced shift in the diffraction profile to higher angles (Figure 1). Diffusional concentration changes cannot be the reason for such an angular shift, since the formation of a nickel-enriched Ti₃Ni₄ phase leads to the depletion of the B2 matrix in nickel, an increase in its lattice parameter, and a corresponding slight shift of the profile towards low angles. Therefore, the reason for the observed profile shift is the residual tensile stresses in the surface layer of the irradiated sample and the corresponding Poisson’s compression normal to the surface.

The application of QC ECAP also leads to the threefold broadening of the X-ray diffraction profile as compared to that at RT. The application of PDA after QC ECAP leads to an asymmetrical distortion and the additional broadening of the general X-ray profile, indicating an increase in rhombohedral distortion and the amount of the R-phase. The observed unevenness of (330)_R and ($3\overline{3}0)$_R intensities indicates an unequal realization of the variants of the orientational relationship between B2- and R-phase lattices that provide each of these two singlets of the {330}_R family.

As QC ECAP creates a high concentration of lattice defects (dislocations, sub-boundaries, and grain boundaries), which in turn promote R-phase formation, the observed wide peak can be a result of the superposition of B2-austenite (110)_B2 and R-phase (330)_R and ($3\overline{3}0)$_R lines. This assumption is confirmed via the TEM observations presented in Section 3.2. Another obvious reason for the growth of the width is an increase in the defectiveness of the crystal lattice profile after dynamic precipitation deformation and hardening during QC ECAP.

3.2. Transmission Electron Microscopy

The TEM microstructural images obtained after the combination of pre-aging for 1 and 5 h and QC ECAP for three passes at a temperature of 450 °C are shown in Figure 2.

Before deformation, the average size of austenite grains recrystallized as a result of the RT was about 30 µm [25,26]. Pre-aging for 1 and 5 h was accompanied by a different intensity of the precipitation of the Ti₃Ni₄ phase [14,22]. An analysis of Figure 2, which shows bright-field and dark-field images and SAED patterns, allows for the conclusion that the structure is abruptly refined as a result of QC ECAP deformation. In general, the observed patterns after the deformation of samples in different structural states are very similar. The main phases in the samples are B2-austenite and the R-phase, for which the amounts cannot be compared, since their strongest reflexes coincide. No traces of the Ti₃Ni₄ phase could be found in the bright-field and dark-field images, which may have been caused by the ultra-small size of particles in the range of 3–5 nm due to their fragmentation or possible dissolution during severe plastic deformation [20,25]. However, a certain number of Ti₃Ni₄ phase reflexes could be identified in the SAED patterns. It seems that after QC ECAP with pre-aging for 5 h, there are more traces of it. The deformation of samples after pre-aging for 1 h allows the formation of a somewhat more finely dispersed structure with an average size of its elements (grain and subgrain) of 92 nm, as compared to that of samples after pre-aging for 5 h (115 nm). It should be noted that, in itself, obtaining a mixed nanograined and nanosubgrained structure in titanium nickelide bulk samples is an important scientific and technological result [27,28,29]. During pre-aging for 1 h, the growth of Ti₃Ni₄ phase particles proceeds less intensively than that after pre-aging for 5 h [30,31,32]. Thus, during QC ECAP, finer Ni₄Ti₃ particles after pre-aging for 1 h provide higher resistance to grain boundary migration and the formation of a finer matrix structure, while after pre-aging for 5 h coarser and rarer Ti₃Ni₄ particles inhibit matrix structure refinement to a lesser extent.

It should be noted that after QC ECAP, there are regions of a submicron size (Figure 2b,d), where a high density of dislocations is observed that are not rearranged into grain and subgrain boundaries. This is evidence that the processes of dynamic recrystallization have not yet been fully completed. Areas of Ti₃Ni₄ phase precipitation, in turn, also become areas of priority recrystallization. The QC ECAP deformation of an aging nickel-rich alloy is accompanied by the formation of a more finely dispersed matrix structure, as compared to that of an equiatomic TiNi SMA [17]. This can presumably be explained by the precipitation of a Ti₃Ni₄ phase that inhibits softening processes.

3.3. Differential Scanning Calorimetry

Results of the DSC study after pre-aging for 1 and 5 h in combination with QC ECAP and PDA, and after RT are shown in the form of DSC curves in Figure 3. The defined characteristic temperatures of forward and reverse MTs are presented in

Table 1. Characteristic temperatures of MT after RT and QC ECAP in combination with pre-aging for 1 and 5 h and PDA.

	Transformation
Treatment	Forward B2→ B19′						Reverse B19′→B2
	Start		Peak		Finish		Start		Peak		Finish
RT	−12		−28		−45		−17		2		8
	Forward						Reverse
	B2 →R			R→B19′			B19′→R			R→B2
	Start	Peak	Finish	Start	Peak	Finish	Start	Peak	Finish	Start	Peak	Finish
RT + 450 °C/1 h	~50	35	0	−20	−44	−58	0	12	21	21	28	50
RT + 450 °C/5 h	~50	37	30	15	−27	−54	31	40	-	-	45	52
RT + 450 °C/1 h + QCECAP	49	25	−10	−27	−72	<−100	0	8	19	-	29	45
RT + 450 °C/1 h + QCECAP + 430 °C/1 h	48	37	22	−39	−70	−89	14	26	35	35	47	55
RT + 450 °C/5 h + QCECAP	41	22	−4	−39	−68	−95	0	9	21	-	25	46
RT + 450 °C/5 h + QCECAP + 430 °C/1 h	43	39	28	−28	−48	−67	19	29	40	40	43	52

.

After the RT, one-stage forward and reverse MTs with a starting temperature of forward MT, M_s, of −12 °C and a finishing temperature of reverse MT, A_f, of 8 °C are observed that are typical of a hot-deformed hyper-equiatomic TiNi SMA with a coarse-grained structure, and this correlates well with the observation of only a B2 X-ray line in the diffractogram (see Figure 1). Aging for 1 and 5 h leads to the appearance of two-stage forward and reverse MTs which pass through an intermediate R-phase and show noticeable broadening of their temperature ranges. The position of the starting temperature of forward MT changes; it increases to 50 °C after aging for 1 h, while after the RT, M_s is equal to −12 °C. The rise of the starting temperature of the first DSC peak attributed to B2→R transformation is explained by the precipitation of Ti₃Ni₄-phase cohered particles promoting the B2→R transformation [1]. The finishing temperature of reverse MT, A_f, increases noticeably to 50 °C after aging for 1 h, and to 52 °C after aging for 5 h.

The application of ECAP after pre-aging for 1 h leads to a sharp decrease in the finishing temperature of forward MT, M_f, to below −100 °C. After ECAP with pre-aging for 5 h, this temperature is slightly higher (−95 °C). Additionally, the distance between the temperature ranges of B2 →R and R→ B19′ becomes significant. These changes can be explained via deformation hardening accompanied by a sharp increase in the density of lattice defects inhibiting B19′-martensite formation.

PDA at 430 °C for 1 h after QC ECAP leads to an increase in the temperatures of forward and reverse MT and a marked separation of peaks, corresponding to B19′→R and R→B2 transformations. The application of annealing is necessary for the setting of the required operational shape of the smart thermosensitive element, produced from TiNi SMA. It can be seen in Table 1 that the finishing temperature of reverse MT after PDA is in the range of 35–40 °C, which closely resembles the temperature of the human body. This makes it possible to use this alloy after applied heat and thermomechanical treatment for the manufacture of medical devices.

3.4. Mechanical and Functional Properties

Results of tension tests in the form of stress–strain diagrams after the RT and QC ECAP in combination with pre-aging and PDA are shown in Figure 4. Based on the obtained diagrams, the following mechanical characteristics were determined: critical stress for martensite reorientation (“transformation”), yield stress, σ_cr, true (“dislocation”) yield stress, σ_y, ultimate tensile strength, σ_B, the difference between dislocation and transformation yield stresses, Δσ = σ_y − σ_cr, and relative elongation to failure, δ. A unifying table,

Table 2. Mechanical and functional properties of TiNi hyper-equiatomic SMA after RT and QC ECAP in combination with pre-aging for 1 and 5 h and PDA.

Treatment	σcr (MPa)	σy (MPa)	Δσ (MPa)	σB (MPa)	δ (%)	HV (e.g.,)	Ɛrt (%)	Ɛf (%)
RT	505	670	165	795	30	214	4.0	0
RT + 450 °C/5 h	170	945	775	1144	27	260	-
RT + 450 °C/1 h + QCECAP	200	1170	970	1250	45	298	11.1	0.3
RT + 450 °C/1 h + QCECAP + 430 °C/1 h	130	1225	1095	1320	36	298	11.3	0.3
RT + 450 °C/5 h + QCECAP	175	1355	1180	1453	42	285	11.2	0.4
RT + 450 °C/5 h + QCECAP + 430 °C/1 h	120	1410	1290	1562	39	286	11.6	0.5

, was compiled, which also includes hardness values (HV) and the functional properties of total recovery strain (Ɛ_rt) and residual strain (Ɛ_f).

The application of aging for 5 h leads to a considerable increase in strength characteristics as compared to those with a RT. After aging dislocation, yield stress rose from 670 MPa to 945 MPa, while ultimate tensile strength rose from 795 MPa to 1144 MPa. The application of QC ECAP after pre-aging for 1 and 5 h leads to an additional increase in strength characteristics; dislocation yield stress and ultimate tensile strength were distinctly higher after QC ECAP with pre-aging for 5 h (σ_y = 1355 MPa σ_B = 1453 MPa) as compared to those after QC ECAP with pre-aging for 1 h (σ_y = 1170 MPa σ_B = 1250 MPa). This can be explained by the incomplete dissolution of Ti₃Ni₄ particles during QC ECAP, which provides a strengthening effect. After the application of QC ECAP, the value of Δσ visibly increased as compared to that after the RT, especially after the addition of PDA [17]. This provided a positive effect on the values of completely recoverable strain. The highest mechanical properties were obtained after ECAP with pre-aging for 5 h in combination with PDA at 430 °C for 1 h: a dislocation yield stress of 1410 MPa and ultimate tensile strength of 1562 MPa. The application of PDA leads to an increase in mechanical properties due to the additional precipitation of the Ti₃Ni₄ phase.

The results of Vickers hardness tests, as shown in Table 2, revealed that maximum hardness was obtained after QC ECAP with pre-aging for 1 h, at a value of 287 HV. After pre-aging for 5 h, hardness was slightly lower, at a value of 276 HV. It is unusual that after PDA and after QC ECAP at 430 °C for 1 h, there was no sharp increase in hardness values, for samples after pre-aging for both 1 and 5 h, while mechanical properties increased drastically. This may be explained by the change in the temperature range of forward MT. Then, hardness was measured near the peak temperature of the B2→R transformation, and precipitation hardening was compensated for with an increase in the size of the indenter footprint, due to B2→R transformation-induced plasticity and a decreased effect of precipitation hardening of Ti₃Ni₄ particles in a highly defective structure on the hardness values.

The application of QC ECAP leads to a sharp increase in total recoverable strain as compared to that under a RT, from 4 to more than 11%, and this can be explained via significant grain refinement. The residual strain after a strain inducement of about 12% for all studied ECAP regimes was less than 0.5%, and the shape recovery rate was more than 95%. The difference in the values of total recoverable strain between ECAP with pre-aging for 1 and 5 h after a strain inducement of about 12% consists in various ratios of superelastic and shape recovery components due to the difference in phase composition, consisting in a higher amount of Ti₃Ni₄ particles, resulting in additional deformation hardening. PDA after QC ECAP leads to a decrease in superelastic components in shape recovery due to the slight shift in the temperature range of reverse MT towards higher temperatures.

4. Conclusions

The effect of static pre-aging time before deformation in combination with the dynamic aging process during quasi-continuous ECAP, and static aging during post-deformation annealing after QC ECAP, on the microstructure formation and thermomechanical and functional properties of a hyper-equiatomic TiNi alloy was studied. QC ECAP leads to the considerable refinement of structural elements and the formation of a mixed fine-grade B2-austenite structure with an average size of structural elements (grains and subgrains) of 92 nm after pre-aging for 1 h, and of 115 nm after pre-aging for 5 h. After QC ECAP in combination with pre-aging for 1 and 5 h, traces of Ti₃Ni₄ particles were not found against the background of a mixed B2- and R-phase matrix in TEM bright- and dark-field images, and were indicated only in SAED patterns, which were caused by the ultra-small size of Ti₃Ni₄ particles due to their fragmentation or possible dissolution during severe plastic deformation. QC ECAP after pre-aging leads to the considerable improvement of mechanical and functional properties as compared to those under a RT. The best combination of mechanical and functional properties is found after QC ECAP with pre-aging for 5 h and post-deformation aging at 430 °C for 1 h: a dislocation yield stress of 1410 MPa, ultimate tensile strength of 1562 MPa, and total recovery strain of 11.6%. These values are comparable with the best results obtained for hyper-equiatomic nickel titanium alloys of a medical grade and expand opportunities for the application of smart shape memory devices.